Qwen3-8B
Released
2025-08-15
Last refreshed
2026-10-26
Status
Researched 20d ago
Open sourceCommercial use: permittedRAGLong contextClassificationJSON / Tool use
Alibaba releases · 51 in the last 12 months · this family litChangelog →
Specifications
- Family
- Qwen3
- Released
- 2025-08-15
- Context
- 128K
- Parameters
- 8B
- Architecture
- Decoder Only
- Specialization
- general
- Openness
- Open source
- License
- Apache 2.0OSI-approvedPermitted
- Weights
- Available
- Code
- Unknown
- Training
- Pretrained
Created by
Pricing
Output / 1M
$0.138
Input / 1M
$0.035
Cheapest of 2 routes · Novita AI
Created by
Pricing
Output / 1M
$0.138
Input / 1M
$0.035
Cheapest of 2 routes · Novita AI
About
Compact base model in Qwen3 series, supporting 100+ languages with strong reasoning and coding capabilities.
Providers(2)
Compare all 2Compare API pricing across 2 tracked provider routes for input and output tokens, batch, and cached reads when available.
| Provider | Input / 1M | Output / 1M | Route |
|---|---|---|---|
| Novita AI | $0.035 | $0.138 | Serverless |
| Featherless | $0.117 | $0.455 | Serverless |
Deprecated provider routes(2)
| Provider | Status | Catalog ID |
|---|---|---|
| Fireworks AI | Deprecated | accounts/fireworks/models/qwen3-8b |
| OpenRouter | Deprecated | qwen/qwen3-8b |
Available via routers & gateways(1)
Capabilities
Structured Outputs
Benchmark peer barsfor RAG
RULERRank 14 of 31
Benchmark scores(3)
Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
| Benchmark | Score | Version | Evaluation | Source |
|---|---|---|---|---|
| 58.9 | diamondObserved 2026-04-18 | — | ||
| 42.6 | v4Observed 2026-10-26 | Configuration: Qwen3-8B (FC) Harness: BFCL V4 official evaluation (gorilla berkeley-function-call-leaderboard), FC mode Evaluator: Gorilla LLM team (UC Berkeley) Confidence: confirmed Notes: BFCL V4 leaderboard (Last Updated 2026-04-12), entry 'Qwen3-8B (FC)': rank 39, Overall Acc 42.57%; Multi Turn 41.75%, Web Search 12.00%, Memory 14.62%, Non-Live AST 87.58%, Live 80.53%, Irrelevance 79.07%. Retrieved 2026-10-26. | ||
| 77.4 | RULER (128K) (accuracy)Observed 2026-10-26 | — |
Migration checks
No linked migration route is available for this model yet.