Qwen 3.8 Flash
Not tested yet
Language modelAPI only
Qwen 3.8 Flash
Alibaba
- Released
- August 26, 2026
- Data date
- September 8, 2026
- Intelligence
- Overall score
- Speed
- 73.4 tokens/sOutput throughput
- Price
- Input · Output / 1M tokens
- Input
- Text, Image, Video
- Output
- Text
No published category-wide ranking is available for this model.
Entries 1-8 of 8. Page 1 of 1.
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
Artificial Analysis Intelligence Index (Show measurement, test conditions, and source)
- Source value
- 42.239646461458
- Score
- 42.239646461458
- Metric
- points
- Unit
- points
- Direction
- Higher is better
- Comparison cohort
- research:qwen-3.8-flash:0
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
AA-Omniscience Index (Show measurement, test conditions, and source)
- Source value
- -9.71666666666667
- Score
- -9.71666666666667
- Metric
- points
- Unit
- points
- Direction
- Higher is better
- Comparison cohort
- research:qwen-3.8-flash:1
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
Output speed (Show measurement, test conditions, and source)
- Source value
- 73.400238403134
- Score
- 73.400238403134
- Metric
- tokens/s
- Unit
- tokens/s
- Direction
- Higher is better
- Comparison cohort
- research:qwen-3.8-flash:2
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
Latency reasoning time (Show measurement, test conditions, and source)
- Source value
- 27.247867902219312
- Score
- 27.247867902219312
- Metric
- seconds
- Unit
- seconds
- Direction
- Lower is better
- Comparison cohort
- research:qwen-3.8-flash:3
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
Latency input time (Show measurement, test conditions, and source)
- Source value
- 2.67264203400003
- Score
- 2.67264203400003
- Metric
- seconds
- Unit
- seconds
- Direction
- Lower is better
- Comparison cohort
- research:qwen-3.8-flash:4
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
Answer generation time (end-to-end timing breakdown) (Show measurement, test conditions, and source)
- Source value
- 6.811966975554828
- Score
- 6.811966975554828
- Metric
- seconds
- Unit
- seconds
- Direction
- Lower is better
- Comparison cohort
- research:qwen-3.8-flash:5
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
End-to-end reasoning time component (Show measurement, test conditions, and source)
- Source value
- 27.247867902219312
- Score
- 27.247867902219312
- Metric
- seconds
- Unit
- seconds
- Direction
- Lower is better
- Comparison cohort
- research:qwen-3.8-flash:6
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
End-to-end input time component (Show measurement, test conditions, and source)
- Source value
- 2.67264203400003
- Score
- 2.67264203400003
- Metric
- seconds
- Unit
- seconds
- Direction
- Lower is better
- Comparison cohort
- research:qwen-3.8-flash:7
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-09-08
- Methodology
- Independent evaluation. unknown
- Context
- Artificial Analysis performance timing; output speed is decoded after the first streaming chunk. Provider basis: first-party API.
Our model tests
Community tests you can inspect
Compare how the models respond to the same prompt. Each test shows the first attempt, with no subsequent fixes to the generated code. These results do not contribute to an overall score.
Pelican on a bicycle
A community classic for free-form SVG drawing.
Prompt and test conditions
Generate an SVG of a pelican riding a bicycle
Token limit including reasoning: 8,192 tokens
Task origin (Simon Willison)Profile
Specifications and access
Published information about this model. Existing estimates are explicitly labeled.
- Weights
- ProprietarySource
- Access
- APISource
- Parameters
- 125BSource
- Active parameters
- 6BSource
- Context window
- 1,000,000 tokensSource
- Notes
- Multimodal QwenCloud API model with deep thinking, visual understanding, and 1M-token context. Qwen's official release page lists a 125B main model, additional 51B N-gram embeddings, and 6B active parameters per token for Qwen3.8-Flash-Next; the production Qwen3.8-Flash variant is served on QwenCloud.Source
Additional source observations
Dated documentation and model-card observations. Provider limits, native context and extended context can differ. Configuration notes retain the source wording.
- API model ID
- qwen3.8-flashSource
- API interfaces
- QwenCloud OpenAI Chat Completions, QwenCloud OpenAI Responses, QwenCloud AnthropicSource
- Input
- Text, Image, VideoSource
- Output
- TextSource
- Reasoning mode
- thinking by default; controllableSource
- Tool calling
- YesSource
- Structured output
- YesSource
- Maximum context window
- 1,000,000 tokensSource
- Maximum output
- 65,536 tokensSource
- Release
- 2026-08-26Source
- Parameters
- 125 BSource
- Active parameters
- 6 BSource
- Maximum images
- 2,048Source
- Maximum videos
- 64Source
- Video limit
- 2 hours / 2 GBSource
- Architecture
- hybrid Qwen3.5-style MoE with n-gram embedding module; 125B main model plus 51B N-gram embeddings and 6B active parameters per tokenSource
- Native context window
- 262,144 tokens (API context is 1M by default; native/base context is separately recorded.)Source
- Source pricing conditions
- $0.15 / 1M input tokens, $0.47 / 1M output tokens (QwenCloud launch-post tariff, retained as a dated observation. The canonical price index separately uses the Alibaba Cloud Global deployment tariff; these are different documented price sources.)Source
Pricing
Published prices
Prices apply to the stated unit. Resolution, output length, and provider can change the cost.
Cost example
100 API requests
$0.0304
100 requests with 1,000 input and 500 output tokens each calculate to $0.0304. Input accounts for $0.0113; output accounts for $0.0191.
- Input price
- $0.113 / 1M tokensBase tier for each request
- Output price
- $0.382 / 1M tokensBase tier for each request
The calculation uses documented token prices and no cache discount. It excludes extra tools, tax, reasoning tokens, and further hidden output tokens.
Open sourceHead-to-head comparisons
Compare this model
Each matchup compares this model with exactly one other model from the same category.
There are no published direct comparisons for this model yet.
More models
Models from the same selection
All AI models
Evidence
Sources and data date
Every statement links to its underlying documentation or leaderboard.
- Speed
Alibaba. Output speed after the first streaming chunk, not total response time. Retrieved September 8, 2026.
- Price
Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs.
- Input
Documented modalities. Retrieved September 8, 2026.
- Output
Documented modalities. Retrieved September 8, 2026.
Research date September 8, 2026. 4 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Open evidence questions and research notes (original wording)
Not found in the inspected sources
knowledge-cutoff. Checked Model Studio model/API pages do not state an exact knowledge cutoff.
Not found in the inspected sources
profile-benchmarks. No exact-model numeric benchmark table was found in the checked API documentation.