Agnes 3.0 Flash vs Claude Fable 5
Agnes 3.0 Flash 
Claude Fable 5 Verdict
- Claude Fable 5 (thinking) leads on quality: 70.9 vs 62.3.
- Agnes 3.0 Flash is stronger in agents & tools.
- Claude Fable 5 (thinking) is stronger in composite, knowledge, long context, reasoning, instruction following.
- Agnes 3.0 Flash is 267× cheaper ($0.075 vs $20.00 per 1M blended).
| Metric | Agnes 3.0 Flash | Claude Fable 5 (thinking) |
|---|---|---|
| BenchLeader Index | 62.3 | 70.9 |
| Agents & tools score | 76.8 | 74.1 |
| Composite score | 74.0 | 92.1 |
| Knowledge score | 59.1 | 84.9 |
| Long context score | 67.0 | 67.7 |
| Reasoning score | 71.1 | 90.8 |
| Instruction following score | – | 63.0 |
| Blended price $/M | $0.075 | $20.00 |
| Output speed | – | 57 tok/s |
| Time to first answer | – | 108.4 s |
| Context window | 1M | 1M |
| AA Intelligence Index | 35.5 | 49.7 |
| IFBench | – | 63.5% |
| AA-LCR | 81.0% | 82.3% |
| AA-Omniscience | -10.6 | 43.3 |
| Terminal-Bench Hard | – | 62.9% |
| GPQA Diamond (AA) | 92.4% | 92.6% |
| Humanity's Last Exam (AA) | 38.5% | 55.5% |
| SciCode (AA) | 51.6% | 61.0% |
| τ²-Bench Telecom (AA) | – | 98.5% |
| Kagi LLM Benchmark | – | 91.4% |
| CritPt | 15.1% | 28.6% |
| GDPval (AA) | 53.7% | 56.6% |
| τ²-Bench Banking (AA) | 47.6% | 38.1% |
| Analyst Agent (AA) | – | 48.8% |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Agnes 3.0 Flash vs Claude Fable 5: questions
- Is Agnes 3.0 Flash better than Claude Fable 5?
- Claude Fable 5 (thinking) leads on quality: 70.9 vs 62.3. The BenchLeader Index combines every independent quality benchmark; Claude Fable 5 (thinking) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Agnes 3.0 Flash better than Claude Fable 5 for agentic tasks?
- Agnes 3.0 Flash scores higher in agentic tasks (77 vs 74 on the category index, where 50 is average).
- Which is cheaper, Agnes 3.0 Flash or Claude Fable 5?
- Agnes 3.0 Flash is cheaper: $0.075 against $20.00 per million tokens, blended at three input tokens per output token.
- Which has the larger context window?
- Both accept 1M tokens of context.