Last verified: August 23, 2026. This independent Qwen model reference separates hosted API products, downloadable open-weight checkpoints, and provider-specific deployments. Qwen is a model ecosystem, not one permanent chatbot or one interchangeable model ID.
Verification status: documentation-verified, not benchmark-tested. Model identity, access, modalities, context limits, licenses, and lifecycle notices were checked against current Qwen, QwenCloud, Alibaba Cloud Model Studio, Hugging Face, GitHub, and Fireworks sources. We did not run a controlled cross-model benchmark for this page. The available Alibaba Cloud account did not provide a working inference path during this review, so no Alibaba output or performance result is claimed.
Quick answer · Hosted models · Open weights · Family map · How to choose · Lifecycle · Methodology · FAQ
There is no single Qwen model list because hosted service IDs and downloadable checkpoints have different access, licensing, pricing, and lifecycle rules:
qwen3.8-max, released August 3, 2026. QwenCloud documents it as a production 2.4T Mixture-of-Experts vision-language model with hybrid thinking and a 1M context category.qwen3.8-max-preview has ended its preview period. QwenCloud still accepts the old string but automatically routes requests to qwen3.8-max and recommends updating configurations to the production ID.qwen3.7-plus and qwen3.7-flash, where the selected route and region support them.qwen3.7-max remains listed where supported, but QwenCloud now recommends qwen3.8-max for the highest-capability tier. No reviewed notice gives Qwen3.7 Max a retirement date.Qwen/Qwen3.8-27B, released August 14, 2026. It is a dense 27B native vision-language model for text, image, and video input with text output, hybrid thinking enabled by default but switchable per request, 262,144 native context with documented extension to 1,000,000 tokens, and an Apache-2.0 license. The larger Qwen/Qwen3.8-2.4T-A95B, released August 12, remains the Max-class open model: it is text-only, thinking-only, has 2.4T total parameters with 95B active parameters, and uses the custom Qwen3.8-Max License.| Exact ID | Status and access | Verified interface | Practical starting point |
|---|---|---|---|
qwen3.8-max |
Current production QwenCloud flagship; also present in Token Plan allowlists | Text, image, and video input; text output; hybrid thinking; 1M context; Function Calling, structured output, and built-in tools listed | Strongest hosted Qwen reasoning tier. Read the Qwen3.8 Max guide. |
qwen3.8-2.4t-a95b |
Hosted by QwenCloud since August 13, 2026 | Text input and output; 1M context; hosted route for the open-series 2.4T model | Max-scale hosted open-series route. Do not substitute the uppercase repository ID. |
qwen3.8-27b |
Hosted by QwenCloud since August 19, 2026 | Text, image, and video input; text output; 1M context; hybrid thinking; Function Calling, structured output, cache, and built-in tools listed | Current lower-cost hosted Qwen3.8 route; verify the exact account and protocol. |
qwen3.8-max-preview |
Retired Token Plan model ID; old calls are automatically routed to qwen3.8-max |
No longer a separate preview runtime; Credits and usage statistics are calculated as production | Update configurations to qwen3.8-max and do not use the retired ID for a separate evaluation. |
qwen3.7-max |
Older hosted Max alias, still listed where supported; no retirement notice verified | 1M context and hosted tools on documented routes | Existing integrations or comparison against the new flagship. See the Qwen3.7 Max guide. |
qwen3.7-plus |
Current balanced hosted alias | 1M context, supported visual input, reasoning, Function Calling, structured output, and built-in tools | Balanced hosted applications where measured quality meets requirements. |
qwen3.7-flash |
Current lower-cost hosted alias | 1M context and broad hosted feature coverage | Lower-cost or lower-latency work after task-specific evaluation. |
qwen3.8-max is the production flagship service ID. QwenCloud also documents the hosted open-series IDs qwen3.8-2.4t-a95b and qwen3.8-27b. The retired qwen3.8-max-preview string automatically routes to production Max on the documented Token Plan route. Qwen separately publishes uppercase repository IDs for the downloadable 2.4T and 27B checkpoints. Hosted IDs, repository IDs, licenses, and feature surfaces are related but not interchangeable. Token Plan is restricted to permitted interactive use in compatible programming and agent tools; use pay as you go for an application backend. See the Qwen API guide and Qwen API pricing guide.
Open-weight models publish downloadable checkpoints. They require compatible software and sufficient hardware, and the exact repository license still needs to be read before deployment.
| Official checkpoint or family | Architecture and modalities | Context and license | Use |
|---|---|---|---|
Qwen/Qwen3.8-27B |
Dense 27B native vision-language post-trained model; text, image, and video input with text output; hybrid thinking is enabled by default and can be disabled | 262,144 native context, extensible to 1,000,000; Apache-2.0 | Latest Qwen open-weight release by date and the current smaller multimodal Qwen3.8 option. A hosted QwenCloud route is documented separately as qwen3.8-27b. |
Qwen/Qwen3.8-2.4T-A95B |
2.4T-total, 95B-active Mixture-of-Experts post-trained text model; thinking is mandatory; official FP8 variant also available | 262,144 native context, extensible to 1,010,000; custom Qwen3.8-Max License | Max-class Qwen3.8 open-weight release; plan for exceptionally large infrastructure and do not assume the hosted service’s vision, non-thinking, or built-in-tool features. It remains larger than Qwen3.8-27B, but it is not the newest checkpoint by release date. |
Qwen/Qwen3.6-35B-A3B |
Mixture-of-experts; about 35B total and 3B active parameters; text, image, and video understanding with text output | 262,144 tokens natively; the model card documents optional extension to about 1.01M with additional configuration and memory; Apache-2.0 | Earlier multimodal open-weight option when compatibility with its MoE architecture is required; do not present it as the newest smaller Qwen model. |
Qwen/Qwen3.6-27B |
Dense general model with native multimodal understanding and hybrid thinking | 262,144-token native context; Apache-2.0 | Earlier dense multimodal checkpoint. For a current 27B deployment, compare it directly with Qwen3.8-27B rather than presenting Qwen3.6-27B as the latest dense option. |
Qwen3.5: 397B-A17B, 122B-A10B, 35B-A3B, 27B, 9B, 4B, 2B, 0.8B |
General multimodal family with dense and MoE sizes and hybrid thinking | Check each official model card; official checkpoints are published under Apache-2.0 | Choose when the broader size range better matches the deployment than the two current Qwen3.6 checkpoints. |
For official files and safer installation paths, use the Qwen download and local setup guide. A hosted alias such as Plus, Max, or Flash is not automatically an open-weight checkpoint.
| Family | Primary job | Current reference points | Important boundary |
|---|---|---|---|
| General and agents | Reasoning, writing, coding, tools, and multimodal understanding | Qwen3.8 open weights (Qwen/Qwen3.8-27B and Qwen/Qwen3.8-2.4T-A95B); hosted qwen3.8-max, qwen3.8-27b, and Qwen3.7 tiers |
Do not mix a hosted service ID with a downloadable checkpoint. |
| Coding | Repository work, code generation, repair, and coding agents | Qwen/Qwen3-Coder-Next, Qwen/Qwen3-Coder-480B-A35B-Instruct, Qwen/Qwen3-Coder-30B-A3B-Instruct |
These specialist checkpoints are non-thinking models; do not claim hidden reasoning output. Open the coding model guide. |
| Vision-language | Image and video understanding with text output | Qwen3.8-27B; Qwen3.6/3.5 general multimodal models; previous dedicated Qwen3-VL Instruct and Thinking checkpoints | Qwen3-VL remains useful for compatibility, but it is not newer than Qwen3.6. Open the Qwen3-VL guide. |
| Omni and audio | Speech recognition, speech generation, and multimodal conversation | Hosted Qwen3.5 Omni and Qwen Audio 3.0, including qwen-audio-3.0-realtime-plus and qwen-audio-3.0-realtime-flash released August 10, 2026; open Qwen3-Omni, Qwen3-ASR, and Qwen3-TTS |
Check whether the exact checkpoint returns text, audio, or both. Open the Qwen3-Omni guide. |
| Image generation | Generate and edit images | Hosted qwen-image-3.0-pro and qwen-image-3.0, released August 4, 2026; earlier Qwen Image 2.0 services; open Qwen-Image checkpoints |
Visual understanding and image generation are different tasks; do not infer public weights from a hosted product name. |
| Embeddings and reranking | Search, retrieval, RAG, and ranking | Hosted qwen3.7-text-embedding, released August 11, 2026; Qwen3 Embedding/Reranker in 0.6B, 4B, and 8B; Qwen3-VL Embedding/Reranker in 2B and 8B |
These are retrieval models, not chat assistants. |
| Historical and compatibility | Existing deployments, research, and migration | Qwen3, Qwen3-Next, Qwen2.5, QwQ, Qwen2, and earlier specialist families | Keep using a working checkpoint when justified, but do not label it “latest.” See the Qwen3, Qwen3-Next, and Qwen2.5 guides. |
Trace the model lineage and hosted-tier boundaries in the Qwen2 guide, Qwen Max guide, Qwen Plus guide, and original Qwen-VL guide. These references are retained for history, compatibility, and migration rather than being presented as the newest generation.
Qwen/Qwen3.8-27B identifies a specific official Hugging Face repository. It is separate from the lowercase hosted ID qwen3.8-27b.qwen3.7-plus identifies a provider service model. It cannot be substituted automatically for a Hugging Face name.35B-A3B, 35B is the approximate total parameter count and A3B is the approximate active parameter count per token.| Requirement | Starting point | Check before committing |
|---|---|---|
| Highest-capability hosted reasoning and multimodal work | qwen3.8-max |
Exact route, region, protocol, PAYG price, context allocation, modality, tools, retention terms, and a representative evaluation. |
| Existing configuration that still uses the retired preview string | Update it to qwen3.8-max; the old ID currently routes to production but should not be treated as a separate model |
Compatibility-route behavior, plan Credits, permitted-use restriction, and a tested production configuration. |
| Balanced hosted applications | qwen3.7-plus |
Measured task quality, region, cost tier, visual/tool requirements, and current lifecycle. |
| Lower-cost hosted work | qwen3.7-flash |
Quality, latency, context tier, tool support, and route-specific price. |
| Max-class self-hosted Qwen | Qwen/Qwen3.8-2.4T-A95B or its official FP8 variant |
Custom license, exceptionally large infrastructure requirements, text-only input, mandatory thinking, serving compatibility, context configuration, and measured throughput. |
| Smaller multimodal self-hosted Qwen | Qwen/Qwen3.8-27B or its official FP8 variant |
Dense 27B architecture; text, image, and video understanding with text output; hybrid thinking; 262,144 native context with optional 1M extension; Apache-2.0; hardware, quantization, serving support, and measured throughput. |
If the decision is between model ecosystems rather than Qwen IDs, use our Qwen vs Google Gemini comparison for managed multimodal platforms and our Qwen vs DeepSeek comparison for open-weight and deployment trade-offs.
After selecting a lane, continue with the Qwen API guide, Qwen API pricing guide, download and local setup guide, or Qwen cloud hosting guide.
qwen3.8-max is the current hosted flagship announced on August 3, 2026.qwen3.8-max-preview has ended its preview period. QwenCloud states that calls using the old ID are automatically routed to qwen3.8-max, with Credits and usage statistics calculated as production. Update the configured model ID.qwen3.7-max is no longer the current flagship recommendation, but no reviewed lifecycle notice announces its retirement.qwen3.8-max-preview string on Token Plan. Do not generalize it to another alias or provider route.qwen3.6-max-preview, qwen3-max, qwen3-vl-flash, and qwen3-coder-plus, with newer replacements listed in its lifecycle notice.Individual model articles listed below can remain useful for compatibility and history. Treat an older guide without a current verification date as pending re-verification, not as an equal recommendation beside the current tables above.
Not tested for this update: comparative quality, throughput, latency, long-context recall, tool-call accuracy, video understanding, speech quality, image generation, safety behavior, regional availability, or production rate limits. No claim on this page should be read as a guarantee for those properties.
As verified on August 23, 2026, qwen3.8-max is QwenCloud’s current production hosted flagship. QwenCloud also documents hosted qwen3.8-27b, launched August 19. By official open-weight release date, Qwen/Qwen3.8-27B is the newest checkpoint, released August 14, two days after Qwen/Qwen3.8-2.4T-A95B. The 27B model is a smaller multimodal Apache-2.0 checkpoint; the 2.4T-A95B model is the larger Max-class text-only, thinking-only checkpoint under the custom Qwen3.8-Max License. Hosted service IDs and repository IDs remain different identifiers.
There is no universal best model. Choose using your required modality, access type, hardware, region, context, cost, latency, license, and a repeatable evaluation on your own workload.
No. Those names normally identify hosted service tiers. Only claim open weights when an exact first-party repository publishes the checkpoint and its license.
The family is related, but the service contracts are not identical. Alibaba/QwenCloud and Fireworks use different exact IDs and document different context and modality limits. Configure and cite the selected provider.
No. A documented maximum is an interface limit, not a guarantee of recall quality, latency, cost, or hardware feasibility. Test the actual workload at the intended length and endpoint.
Use What Is Qwen AI? for ownership, history, terminology, and the wider ecosystem. Use this Models hub for current families, exact identifiers, access, status, and selection.
The archive below contains individual model and family guides. Start with the current tables above, then open the exact guide you need. Older entries are retained for historical and migration value and should be checked against their visible verification date and official sources.
Independent, source-linked dataset
Search exact hosted IDs, provider routes, open repositories, lifecycle states, retirement dates, and documented replacements. A record is unique to its provider and deployment scope; the same ID can appear more than once when official routes differ.
Verified 2026-08-23: Curated, not exhaustive. This dataset covers the Qwen identifiers most relevant to the site's Models cluster, including current QwenCloud production IDs, selected previews and legacy IDs, hosted IDs with announced lifecycle events, representative official open-weight repositories, and one third-party route. Status, limits, replacement guidance, and pricing boundaries are scoped to the cited provider and product route. A hosted-service retirement must not be applied to an open repository with a similar name. Blank values mean that this verification pass did not establish the fact from the cited source; they are not zeroes. No authenticated-console inference or temporary promotional price is treated as a universal fact.
52 records shown.
No records match all selected filters. Reset the filters or check the exact model ID.
Qwen/Qwen3-VL-7B-Instruct
The official Qwen3-VL lineup documents an 8B tier, not a 7B tier. Preserve any public correction/FAQ, but do not emit a valid-model record for the 7B string. Suggested official ID: Qwen/Qwen3-VL-8B-Instruct.
Official correction source.
qwen3.6-flash
Qwen officially relates the hosted service to the open model, but the API ID and repository ID are not interchangeable. Suggested official ID: qwen3.6-flash, Qwen/Qwen3.6-35B-A3B.
Official correction source.
qwen3.7-plus
Use qwen3.7-plus on QwenCloud and the fully qualified accounts/fireworks/models/qwen3p7-plus on Fireworks. Suggested official ID: qwen3.7-plus, accounts/fireworks/models/qwen3p7-plus.
Official correction source.
qwen3.8-max-preview
The preview ID is retired. On the documented Token Plan route, legacy calls are temporarily accepted and automatically routed to production qwen3.8-max. Update the configured ID. Suggested official ID: qwen3.8-max.
Official correction source.
qwen2.5-omni-7b
The hosted China route has a scheduled retirement, the QwenCloud international route is listed as legacy in this verification pass, and the official open repository remains published. Suggested official ID: qwen2.5-omni-7b, Qwen/Qwen2.5-Omni-7B.
Official correction source.
| Model and exact ID | Provider and route | Status | Lifecycle and replacement | Route-bound limits | Open-weight relation | Verification |
|---|---|---|---|---|---|---|
Qwen3.8 Max
qwen3.8-max
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud vision-model matrix route: context 1M / max output 128K; QwenCloud marketplace model card: max input 991K / max output 131K |
No official repository mapping stated Production ID released on 2026-08-03; do not substitute the older qwen3.8-max-preview plan ID. |
QwenCloud model release changelog
Checked 2026-08-04 Internal guide |
Qwen3.8 Max Previewqwen3.8-max-previewHosted Api | QwenCloud QwenCloud Token Plan international plan-specific hosted API | Retired | No retirement date stated Replacement: qwen3.8-max The retired input string is temporarily accepted and automatically routed to production qwen3.8-max on the documented Token Plan route. | See the linked route-specific source | No official repository mapping stated Retired compatibility input string, not a separate live preview runtime. Update configurations to qwen3.8-max. | QwenCloud Token Plan Checked 2026-08-23 Internal guide 2 official sources |
Qwen3.7 Max
qwen3.7-max
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 1M / max output 64K; QwenCloud marketplace model card: max output 131K |
No official repository mapping stated Current production model, but no longer the newest overall after qwen3.8-max; no retirement notice was found. |
QwenCloud model selection
Checked 2026-08-04 Internal guide |
Qwen3.7 Plus
qwen3.7-plus
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 1M / max output 64K; QwenCloud marketplace model card: max output 131K |
No official repository mapping stated QwenCloud ID; it is not interchangeable with Fireworks accounts/fireworks/models/qwen3p7-plus. |
QwenCloud model selection
Checked 2026-08-04 Internal guide |
Qwen3.7 Flash
qwen3.7-flash
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 1M / max output 64K; QwenCloud marketplace model card: max output 131K |
No official repository mapping stated Current QwenCloud speed/cost-oriented production route. |
QwenCloud model selection
Checked 2026-08-04 Internal guide |
Qwen3.6 Plus
qwen3.6-plus
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 1M / max output 64K |
No official repository mapping stated Available production ID; not described as retired. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3.6 Flash
qwen3.6-flash
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud hosted qwen3.6-flash route: context 1M / max output 64K |
Qwen/Qwen3.6-35B-A3B Qwen officially links this hosted service to Qwen/Qwen3.6-35B-A3B, but the API ID and repository ID are different identifiers with different deployment semantics. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide |
Qwen3.6 35B A3B
qwen3.6-35b-a3b
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 256K / max output 64K; QwenCloud vision-model matrix: context 32K / max output 8K |
Qwen/Qwen3.6-35B-A3B Hosted API identifier and official open repository are related but not interchangeable. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide |
Qwen3.6 27B
qwen3.6-27b
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Production |
No retirement date stated Replacement: No official replacement stated |
QwenCloud text-model matrix: context 256K / max output 64K; QwenCloud vision-model matrix: context 32K / max output 8K |
Qwen/Qwen3.6-27B Hosted API identifier and official open repository are related but not interchangeable. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide 3 official sources |
Qwen3.6 Max Preview
qwen3.6-max-preview
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-max QwenCloud names qwen3.7-max as the replacement. |
See the linked route-specific source |
No official repository mapping stated Lifecycle applies only to this QwenCloud service ID. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 Max
qwen3-max
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-max QwenCloud names qwen3.7-max as the replacement. |
See the linked route-specific source |
No official repository mapping stated Do not apply this notice to qwen-max or qwen3.7-max; exact IDs matter. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 Coder Plus
qwen3-coder-plus
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-plus QwenCloud names qwen3.7-plus as the replacement. |
See the linked route-specific source |
No official repository mapping stated Retirement applies to the hosted ID, not to open Qwen3-Coder repositories. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 Coder Next
qwen3-coder-next
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-plus QwenCloud names qwen3.7-plus as the replacement. |
See the linked route-specific source |
Qwen/Qwen3-Coder-Next Hosted qwen3-coder-next is scheduled for retirement; Qwen/Qwen3-Coder-Next remains a separately published repository. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 Coder 480B A35B Instruct
qwen3-coder-480b-a35b-instruct
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-plus QwenCloud names qwen3.7-plus as the replacement. |
See the linked route-specific source |
Qwen/Qwen3-Coder-480B-A35B-Instruct Hosted lifecycle does not unpublish the open repository. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 Coder 30B A3B Instruct
qwen3-coder-30b-a3b-instruct
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-plus QwenCloud names qwen3.7-plus as the replacement. |
See the linked route-specific source |
Qwen/Qwen3-Coder-30B-A3B-Instruct Hosted lifecycle does not unpublish the open repository. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 VL Flash
qwen3-vl-flash
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-flash Current QwenCloud deprecation table names qwen3.7-flash as the replacement. |
See the linked route-specific source |
No official repository mapping stated The replacement is captured from the current QwenCloud route; other Alibaba documentation routes may show different migration wording. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 VL 8B Instruct
qwen3-vl-8b-instruct
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-flash QwenCloud names qwen3.7-flash as the replacement for this hosted route. |
See the linked route-specific source |
Qwen/Qwen3-VL-8B-Instruct Hosted retirement does not retire the official open repository. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 8B
qwen3-8b
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-flash QwenCloud names qwen3.7-flash as the replacement. |
See the linked route-specific source |
No official repository mapping stated This row asserts only the hosted service lifecycle; no repository mapping is asserted in this curated pass. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 14B
qwen3-14b
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-flash QwenCloud names qwen3.7-flash as the replacement. |
See the linked route-specific source |
No official repository mapping stated This row asserts only the hosted service lifecycle; no repository mapping is asserted in this curated pass. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide |
Qwen3 Next 80B A3B Instruct
qwen3-next-80b-a3b-instruct
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Scheduled Retirement |
2026-10-10 Replacement: qwen3.7-plus QwenCloud names qwen3.7-plus as the replacement. |
See the linked route-specific source |
Qwen/Qwen3-Next-80B-A3B-Instruct Hosted retirement does not unpublish the official open repository. |
QwenCloud model deprecation notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen Turbo
qwen-turbo
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Legacy |
No retirement date stated Replacement: No official replacement stated No QwenCloud international retirement date was established from the cited QwenCloud routes. |
QwenCloud international legacy matrix: context 128K / max output 16K |
No official repository mapping stated Do not import the China Model Studio retirement date into this international route. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide 2 official sources |
Qwen Flash
qwen-flash
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Legacy |
No retirement date stated Replacement: No official replacement stated |
QwenCloud international legacy matrix: context 1M / max output 32K |
No official repository mapping stated Legacy availability is not equivalent to a published retirement date. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide |
Qwen2.5 Omni 7B
qwen2.5-omni-7b
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Legacy |
No retirement date stated Replacement: No official replacement stated No QwenCloud international retirement date was established from the cited QwenCloud routes. |
QwenCloud text-model matrix: context 32K / max output 8K; QwenCloud vision-model matrix: context 32K / max output 2K |
Qwen/Qwen2.5-Omni-7B China Model Studio schedules the hosted ID for retirement, while this international route and the open repository require separate status records. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide |
Qwen Max
qwen-max
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Legacy |
No retirement date stated Replacement: No official replacement stated No retirement notice for this exact ID was established in the cited QwenCloud routes. |
QwenCloud international legacy matrix: context 32K / max output 8K |
No official repository mapping stated Do not confuse qwen-max with qwen3-max, which has a separate scheduled-retirement record. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide 2 official sources |
Qwen Plus
qwen-plus
Hosted Api
|
QwenCloud QwenCloud PAYG API international hosted API |
Legacy |
No retirement date stated Replacement: No official replacement stated |
QwenCloud international legacy matrix: context 1M / max output 32K |
No official repository mapping stated Legacy availability is not equivalent to a published retirement date. |
QwenCloud text generation model matrix
Checked 2026-08-04 Internal guide |
Qwen Turbo
qwen-turbo
Hosted Api
|
Alibaba Cloud Model Studio Bailian / Model Studio China China hosted API |
Scheduled Retirement |
2026-10-10T00:00:00+08:00 Replacement: No official replacement stated Official guidance points to the latest models in the Qwen3.7 and Qwen3.6 series; it does not name one exact replacement ID. |
See the linked route-specific source |
No official repository mapping stated This retirement is route-specific and must not be copied onto QwenCloud international without its own notice. |
Alibaba Cloud Model Studio model retirement notices
Checked 2026-08-04 Internal guide |
Qwen Turbo Realtime
qwen-turbo-realtime
Hosted Api
|
Alibaba Cloud Model Studio Bailian / Model Studio China China hosted API |
Scheduled Retirement |
2026-10-10T00:00:00+08:00 Replacement: No official replacement stated Official guidance points to the latest models in the Qwen3.7 and Qwen3.6 series; it does not name one exact replacement ID. |
See the linked route-specific source |
No official repository mapping stated This retirement is route-specific; no equivalent QwenCloud international record is asserted here. |
Alibaba Cloud Model Studio model retirement notices
Checked 2026-08-04 Internal guide |
Qwen2.5 Omni 7B
qwen2.5-omni-7b
Hosted Api
|
Alibaba Cloud Model Studio Bailian / Model Studio China China hosted API |
Scheduled Retirement |
2026-10-10T00:00:00+08:00 Replacement: qwen3.5-omni-plus China Model Studio names qwen3.5-omni-plus as the replacement. |
See the linked route-specific source |
Qwen/Qwen2.5-Omni-7B The scheduled retirement affects this hosted China route, not Qwen/Qwen2.5-Omni-7B open weights. |
Alibaba Cloud Model Studio model retirement notices
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3.7 Plus on Fireworks
accounts/fireworks/models/qwen3p7-plus
Third Party
|
Fireworks AI Fireworks Serverless API third-party hosted API |
Available |
No retirement date stated Replacement: No official replacement stated |
Fireworks serverless route: context 262K |
No official repository mapping stated This provider-qualified ID is not interchangeable with QwenCloud qwen3.7-plus. |
Fireworks AI Qwen3.7 Plus launch
Checked 2026-08-04 Internal guide |
Qwen3.6 35B A3B Open Weights
Qwen/Qwen3.6-35B-A3B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository configuration and memory permit: context 262,144 native; 1,010,000 with documented extension |
Qwen/Qwen3.6-35B-A3B Officially associated with hosted qwen3.6-flash, but this repository has its own context configuration and deployment cost. |
Official Qwen3.6 35B A3B release
Checked 2026-08-04 Internal guide |
Qwen3.6 27B Open Weights
Qwen/Qwen3.6-27B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository configuration and memory permit: context 262,144 native; 1,010,000 with documented extension |
Qwen/Qwen3.6-27B Open repository and hosted qwen3.6-27b are separate resolver records. |
Qwen/Qwen3.6-27B
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 Coder Next Open Weights
Qwen/Qwen3-Coder-Next
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository model card: context 262,144 native |
Qwen/Qwen3-Coder-Next The hosted qwen3-coder-next retirement does not retire this repository. |
Qwen/Qwen3-Coder-Next
Checked 2026-08-04 Internal guide |
Qwen3 Coder 480B A35B Instruct Open Weights
Qwen/Qwen3-Coder-480B-A35B-Instruct
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Documented self-hosting configuration: context 262,144 native; extendable to 1M with YaRN |
Qwen/Qwen3-Coder-480B-A35B-Instruct Hosted lifecycle and repository availability are separate. |
Qwen/Qwen3-Coder-480B-A35B-Instruct
Checked 2026-08-04 Internal guide |
Qwen3 Coder 30B A3B Instruct Open Weights
Qwen/Qwen3-Coder-30B-A3B-Instruct
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Documented self-hosting configuration: context 262,144 native; extendable to 1M with YaRN |
Qwen/Qwen3-Coder-30B-A3B-Instruct Hosted lifecycle and repository availability are separate. |
Qwen/Qwen3-Coder-30B-A3B-Instruct
Checked 2026-08-04 Internal guide |
Qwen3 VL 8B Instruct Open Weights
Qwen/Qwen3-VL-8B-Instruct
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Documented self-hosting configuration: context 256K native; expandable to 1M |
Qwen/Qwen3-VL-8B-Instruct Official Qwen3-VL open lineup uses 8B; no official Qwen3-VL-7B-Instruct repository was verified. |
Official Qwen3-VL GitHub repository
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 VL 8B Thinking Open Weights
Qwen/Qwen3-VL-8B-Thinking
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Documented self-hosting configuration: context 256K native; expandable to 1M |
Qwen/Qwen3-VL-8B-Thinking Distinct Thinking repository; do not silently substitute the Instruct repository. |
Official Qwen3-VL GitHub repository
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 Omni 30B A3B Instruct Open Weights
Qwen/Qwen3-Omni-30B-A3B-Instruct
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
See the linked route-specific source |
Qwen/Qwen3-Omni-30B-A3B-Instruct Open multimodal repository; no hosted-service status is inferred. |
Qwen/Qwen3-Omni-30B-A3B-Instruct
Checked 2026-08-04 Internal guide |
Qwen2.5 Omni 7B Open Weights
Qwen/Qwen2.5-Omni-7B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
See the linked route-specific source |
Qwen/Qwen2.5-Omni-7B The China-hosted qwen2.5-omni-7b retirement does not affect these published open weights. |
Qwen/Qwen2.5-Omni-7B
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3 ASR 1.7B Open Weights
Qwen/Qwen3-ASR-1.7B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
See the linked route-specific source |
Qwen/Qwen3-ASR-1.7B Repository row only; no hosted ID is inferred. |
Qwen/Qwen3-ASR-1.7B
Checked 2026-08-04 Internal guide |
Qwen3 TTS 12Hz 1.7B CustomVoice Open Weights
Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
See the linked route-specific source |
Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice Repository row only; no hosted ID is inferred. |
Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice
Checked 2026-08-04 Internal guide |
Qwen Image Open Weights
Qwen/Qwen-Image
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
See the linked route-specific source |
Qwen/Qwen-Image Repository row only; no hosted ID is inferred. |
Qwen/Qwen-Image
Checked 2026-08-04 Internal guide |
Qwen3 Embedding 8B Open Weights
Qwen/Qwen3-Embedding-8B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository model card: context 32K |
Qwen/Qwen3-Embedding-8B Embedding repository, not a chat-completion model ID. |
Qwen/Qwen3-Embedding-8B
Checked 2026-08-04 Internal guide |
Qwen3 Reranker 8B Open Weights
Qwen/Qwen3-Reranker-8B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository model card: context 32K |
Qwen/Qwen3-Reranker-8B Reranker repository, not a chat-completion model ID. |
Qwen/Qwen3-Reranker-8B
Checked 2026-08-04 Internal guide |
Qwen3 VL Embedding 8B Open Weights
Qwen/Qwen3-VL-Embedding-8B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository model card: context 32K |
Qwen/Qwen3-VL-Embedding-8B Multimodal embedding repository, not a chat-completion model ID. |
Qwen/Qwen3-VL-Embedding-8B
Checked 2026-08-04 Internal guide |
Qwen3 VL Reranker 8B Open Weights
Qwen/Qwen3-VL-Reranker-8B
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository model card: context 32K |
Qwen/Qwen3-VL-Reranker-8B Multimodal reranker repository, not a chat-completion model ID. |
Qwen/Qwen3-VL-Reranker-8B
Checked 2026-08-04 Internal guide |
Qwen3 Next 80B A3B Instruct Open Weights
Qwen/Qwen3-Next-80B-A3B-Instruct
Repository
|
Qwen Official Qwen repository self-hosted open weights |
Published |
No retirement date stated Replacement: No official replacement stated |
Repository configuration and memory permit: context 262,144 native; 1,010,000 with documented extension |
Qwen/Qwen3-Next-80B-A3B-Instruct The hosted ID's scheduled retirement does not retire this repository. |
Qwen/Qwen3-Next-80B-A3B-Instruct
Checked 2026-08-04 Internal guide 2 official sources |
Qwen3.8 2.4T A95Bqwen3.8-2.4t-a95bHosted Api | QwenCloud QwenCloud PAYG API international hosted API | Production | No retirement date stated Replacement: No official replacement stated | QwenCloud marketplace model card: context 1M / max input 991K / max output 131K | Qwen/Qwen3.8-2.4T-A95B, Qwen/Qwen3.8-2.4T-A95B-FP8 Hosted lowercase API ID and uppercase Qwen repository IDs are separate access routes. The repository license does not replace QwenCloud service terms. | QwenCloud model release changelog Checked 2026-08-23 Internal guide |
Qwen3.8 27Bqwen3.8-27bHosted Api | QwenCloud QwenCloud PAYG API international hosted API | Production | No retirement date stated Replacement: No official replacement stated | QwenCloud marketplace model card: context 1M / max input 991K; 983K in thinking mode / max output 131K | Qwen/Qwen3.8-27B, Qwen/Qwen3.8-27B-FP8 QwenCloud launched this hosted ID on 2026-08-19. It is not interchangeable with the Qwen/Qwen3.8-27B repository ID. | QwenCloud model release changelog Checked 2026-08-23 Internal guide |
Qwen3.8 2.4T A95BQwen/Qwen3.8-2.4T-A95BRepository | Qwen Official Hugging Face repository downloadable open-weight checkpoint | Published | No retirement date stated Replacement: No official replacement stated | Official repository model card: context 262,144 native | Qwen/Qwen3.8-2.4T-A95B, Qwen/Qwen3.8-2.4T-A95B-FP8 2.4T-total, 95B-active text checkpoint; thinking-only; 262,144 native context with documented extension to 1,010,000. Hosted lowercase API IDs remain separate service routes. License: Custom Qwen3.8-Max License | Official Qwen3.8 repository Checked 2026-08-23 Internal guide 2 official sources |
Qwen3.8 2.4T A95B FP8Qwen/Qwen3.8-2.4T-A95B-FP8Repository | Qwen Official Hugging Face repository downloadable open-weight checkpoint | Published | No retirement date stated Replacement: No official replacement stated | Official repository model card: context 262,144 native | Qwen/Qwen3.8-2.4T-A95B-FP8, Qwen/Qwen3.8-2.4T-A95B Official FP8 variant of the 2.4T-total, 95B-active text-only, thinking-only checkpoint. Hosted lowercase API IDs remain separate service routes. License: Custom Qwen3.8-Max License | Official Qwen3.8 repository Checked 2026-08-23 Internal guide 2 official sources |
Qwen3.8 27BQwen/Qwen3.8-27BRepository | Qwen Official Hugging Face repository downloadable open-weight checkpoint | Published | No retirement date stated Replacement: No official replacement stated | Official repository model card: context 262,144 native | Qwen/Qwen3.8-27B, Qwen/Qwen3.8-27B-FP8 Dense 27B native vision-language checkpoint; 262,144 native context with documented extension to 1,000,000; thinking can be disabled. Hosted lowercase API IDs remain separate service routes. License: Apache License 2.0 | Official Qwen3.8 repository Checked 2026-08-23 Internal guide 2 official sources |
Qwen3.8 27B FP8Qwen/Qwen3.8-27B-FP8Repository | Qwen Official Hugging Face repository downloadable open-weight checkpoint | Published | No retirement date stated Replacement: No official replacement stated | Official repository model card: context 262,144 native | Qwen/Qwen3.8-27B-FP8, Qwen/Qwen3.8-27B Official fine-grained FP8 variant of the dense Qwen3.8-27B native vision-language checkpoint. Hosted lowercase API IDs remain separate service routes. License: Apache License 2.0 | Official Qwen3.8 repository Checked 2026-08-23 Internal guide 2 official sources |
| Tier or ID | Use it as | Lifecycle boundary |
|---|---|---|
qwen3.8-max | Current QwenCloud flagship for strongest hosted reasoning | Production ID released August 3, 2026; separate from the preview ID |
qwen3.7-plus | Balanced hosted performance, tools and cost | Current QwenCloud recommendation |
qwen3.7-flash | Lower-cost hosted work after task-specific evaluation | Current QwenCloud recommendation |
qwen-turbo | Migration-only identifier on the affected Alibaba Cloud Model Studio route | Scheduled there for retirement October 10, 2026; choose a current Qwen3.7 or Qwen3.6 route instead |
Boundary: the retirement notice is route-specific. It does not retire unrelated provider deployments or downloadable open weights.
qwen3.6-35b-a3b and qwen3.6-27b in its text and visual matrices. Both route-specific rows are preserved; the Resolver does not invent one universal number.Suggested citation: Qwen-AI.chat. Qwen Model ID & Lifecycle Resolver, version 1.1.0. Model records verified August 23, 2026. https://qwen-ai.chat/models/#qwen-model-resolver. Licensed under CC BY 4.0. Accessed [date].
Dataset license: To the extent Qwen-AI.chat owns copyright or database rights, its original selection, arrangement, normalization, and explanatory annotations in this compiled dataset are licensed under Creative Commons Attribution 4.0 International (CC BY 4.0).
Attribution: Credit Qwen-AI.chat, name the Qwen Model ID & Lifecycle Resolver and dataset version, link to the Resolver page and the CC BY 4.0 license, indicate if changes were made, and do not imply endorsement.
License boundary: The license does not cover external provider documentation, source pages, quoted material, model weights, code, repository assets, model or product names, logos, trademarks, or linked materials. Each records[].license value describes only the exact model or repository license verified at the cited source; null means not verified. Source-registry entries are citations, not grants of permission.
Last verified: August 13, 2026. Independent and unofficial: Qwen-AI.chat is not operated by, affiliated with, or endorsed by Alibaba Cloud, QwenCloud, Alibaba Group, or the Qwen team. The specifications and prices below were checked against first-party pages on the date…
Last verified: August 4, 2026. This independent guide is not operated by or affiliated with Alibaba Cloud, QwenCloud, or the Qwen team. Verify prices, quotas, limits, and regional availability on the exact provider route before production use. Qwen3.6 Flash remains…
Last verified: August 4, 2026. This is an independent technical guide and is not operated by or affiliated with Alibaba Cloud or the Qwen team. Confirm availability, billing, and regional requirements in the provider documentation before deployment. Qwen3.7 Plus is…
Last verified: August 3, 2026. This independent guide is not operated by or affiliated with Alibaba Cloud, QwenCloud, or the Qwen team. Model specifications, availability, retirement notices, and prices should be checked against the linked provider documentation before production use.…
Qwen3-Omni is an open-weight multimodal model family developed by the Qwen team. It can understand text, images, audio, and video, but its outputs depend on the checkpoint you select. The Instruct checkpoint can return text and natural speech, while the…
Qwen3 Max, identified in Alibaba Model Studio as qwen3-max, is a hosted commercial text-generation model with a 256K context window. It is different from qwen3.7-max, which is a separate 1M-context model. Last verified: August 4, 2026. Lifecycle notice: Alibaba schedules…
Last verified: August 4, 2026. Qwen2.5-Max was a proprietary, hosted large-scale Mixture-of-Experts model announced by the Qwen team on January 28, 2025. It was not a downloadable Qwen2.5 checkpoint, and its launch API snapshot should not be used as a…
Qwen2-VL is the Qwen Team’s 2024 vision-language family for text, image, and video understanding with text output. Its official 2B, 7B, and 72B Base and Instruct checkpoints remain publicly downloadable. Qwen2-VL is now a historical compatibility option rather than the…
Qwen3-Next is an open-weight, text-only mixture-of-experts model family. Its public release uses separate Instruct and Thinking checkpoints. It is not one downloadable checkpoint that switches dynamically between the two behaviors. Last verified: August 4, 2026. Checkpoint rule: Select Qwen/Qwen3-Next-80B-A3B-Instruct for…
Qwen3-Coder is a family of text-based coding language models developed by the Qwen team. It includes large mixture-of-experts Instruct checkpoints, the separate Qwen3-Coder-Next architecture, downloadable model variants, and hosted Alibaba Cloud Model Studio IDs. Last verified: August 23, 2026. The…
Verification status Last reviewed and verified: August 3, 2026. What was checked: the official Qwen release article, Qwen’s first-party Hugging Face collection and model cards, generator configuration files, repository licenses, and the current Alibaba Cloud model-lifecycle documentation. Public repository pages…
Qwen3-VL is an open-weight vision-language model family that accepts text, images, and video and produces text. Qwen publishes Dense and mixture-of-experts sizes, with separate Instruct and Thinking checkpoints for every core size. Last verified: August 4, 2026. Model-ID correction: Qwen/Qwen3-VL-7B-Instruct…
Qwen2.5 is a September 2024 generation of dense, decoder-only language models from the Qwen team. The core family contains seven text-model sizes, each published in Base and Instruct forms. Last verified: August 4, 2026. Core Qwen2.5 checkpoints are text-in, text-out…
Qwen3 was released on April 29, 2025 as a family of dense and mixture-of-experts language models. The April release is also identified as Qwen3-2504 to distinguish it from later dated checkpoints. Last verified: August 4, 2026. The official core lineup…
Qwen2.5-Omni is an open-weight multimodal model family from the Qwen team. It can process text, images, audio, and video, then generate a text response and, when the Talker component is enabled, synthesized speech. It does not generate images or videos.…
Verification status Last reviewed and verified: August 3, 2026. What was checked: Qwen’s original Qwen2-Math announcement, the first-party Qwen2-Math collection and model cards, generator configuration files, exact repository licenses, and Alibaba Cloud’s current hosted-model retirement documentation. Public repository pages were…
Qwen2.5-Coder is a 2024 open-weight model family specialized for code generation, code explanation, code repair, and related text-based programming tasks. It is based on Qwen2.5 and is preserved here as a historical coding generation. Last verified: August 4, 2026. Qwen2.5-Coder…
Qwen2 is a 2024 generation of open-weight language models developed by the Qwen team. It includes dense and mixture-of-experts checkpoints in several sizes, but its context limits and licenses are not identical across the family. Last verified: August 10, 2026.…
Qwen-VL is the original 2023 vision-language model family released by the Qwen Team. This page covers two specific public checkpoints: the pretrained Qwen/Qwen-VL model and the aligned Qwen/Qwen-VL-Chat assistant. It is a historical and compatibility guide, not a page for…
Last verified: August 2, 2026. Qwen2.5-VL is a 2025 vision-language model family from the Qwen team. It accepts text, images, and video and generates text. It does not generate images or audio. The family remains useful when you need downloadable…
Qwen Max is a hosted model tier, not one permanent model or a downloadable checkpoint. The current production flagship is qwen3.8-max, launched by QwenCloud on August 3, 2026. qwen3.7-max remains active and retains a separate operational role: it is still…
Accuracy note: QwQ-32B is not documented as an AI companion or character model. The official sources do not attribute persistent memory, an emotional state, a permanent persona, consciousness, or cross-session recall to the checkpoint. An application may supply conversation history,…
Last verified: August 23, 2026. Hosted lifecycle notice (route-specific): Alibaba Cloud’s Bailian / China Model Studio lifecycle page schedules the hosted IDs qwen-turbo and qwen-turbo-realtime for decommissioning on October 10, 2026. Its replacement guidance is the latest Qwen 3.7 and…
Last verified: August 4, 2026. “Qwen Flash” is now a model tier rather than one fixed downloadable checkpoint. The name can refer to a current Qwen3.7 or Qwen3.6 Flash service, or to the older qwen-flash API family. Always record the…
Qwen Plus, identified as qwen-plus in Alibaba Model Studio, is a legacy hosted text-generation model with a 1M context window. It is a commercial API model, not a downloadable checkpoint intended for local deployment. Last verified: July 21, 2026. Important…