Claude Fable 5 vs GPT-5.2-Codex
Claude Fable 5 
GPT-5.2-Codex Verdict
- Claude Fable 5 leads on quality: 70.5 vs 60.8.
- Claude Fable 5 is stronger in agents & tools, coding, composite, human preference, knowledge, multimodal, reasoning.
- GPT-5.2-Codex is stronger in instruction following.
- GPT-5.2-Codex is 4.2× cheaper ($4.81 vs $20.00 per 1M blended).
- Claude Fable 5 streams 2.5× faster (63 vs 25 tokens per second).
| Metric | Claude Fable 5 | GPT-5.2-Codex |
|---|---|---|
| BenchLeader Index | 70.5 | 60.8 |
| Agents & tools score | 74.4 | 63.5 |
| Coding score | 70.5 | 55.7 |
| Composite score | 91.8 | 56.8 |
| Human preference score | 71.6 | – |
| Instruction following score | 62.9 | 75.1 |
| Knowledge score | 71.4 | 63.1 |
| Long context score | 67.8 | 67.8 |
| Multimodal score | 69.7 | 60.7 |
| Reasoning score | 75.9 | – |
| Blended price $/M | $20.00 | $4.81 |
| Output speed | 63 tok/s | 25 tok/s |
| Time to first answer | 102.3 s | 6.7 s |
| Context window | 1M | 400k |
| Terminal-Bench | – | 66.5% |
| Remote Labor Index | 16.1% | – |
| APEX-Agents | 45.0% | 27.6% |
| FrontierCode | 53.5% | – |
| Epoch Capabilities Index | 163.4 | – |
| LMArena Text | 1507 | – |
| LMArena Hard Prompts | 1533 | – |
| LMArena Coding | 1551 | – |
| LMArena WebDev | 1628 | 1339 |
| LMArena Vision | 1330 | – |
| LiveBench | – | 74.0% |
| LiveBench Reasoning | – | 77.7% |
| LiveBench Coding | – | 83.6% |
| LiveBench Agentic Coding | – | 49.4% |
| LiveBench Mathematics | – | 88.8% |
| LiveBench Data Analysis | – | 78.2% |
| LiveBench Language | – | 73.7% |
| AA Intelligence Index | 49.7 | 28.5 |
| IFBench | 63.5% | 77.6% |
| AA-LCR | 82.3% | 82.3% |
| MMMU-Pro | – | 76.3% |
| AA-Omniscience | 43.3 | -2.2 |
| Terminal-Bench Hard | 62.9% | 37.1% |
| GPQA Diamond (AA) | 92.6% | 89.9% |
| Humanity's Last Exam (AA) | 55.5% | 35.7% |
| SciCode (AA) | 61.0% | – |
| τ²-Bench Telecom (AA) | 98.5% | 92.1% |
| LiveCodeBench | 89.8% | 88.0% |
| MMLU-Pro | 91.5% | – |
| IOI | 72.3% | – |
| LegalBench | 88.6% | – |
| CorpFin | 71.8% | – |
| TaxEval | 76.9% | – |
| Terminal-Bench 2.1 (Vals) | 80.5% | – |
| SWE-bench (Vals) | 95.0% | – |
| GPQA Diamond (Vals) | 93.2% | – |
| Vals Index | 66.0 | – |
| SWE-Bench Pro | – | 41.0% |
| MCP Atlas | 83.3% | – |
| PRBench Finance | 53.9% | – |
| PRBench Legal | 52.6% | – |
| HiL-Bench | 56.3% | – |
| EQ-Bench 4 | 1340 | – |
| Kagi LLM Benchmark | 88.8% | – |
| SWE-bench Verified (bash only) | – | 72.8% |
| SWE-bench Verified (any scaffold) | – | 72.8% |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Claude Fable 5 vs GPT-5.2-Codex: questions
- Is Claude Fable 5 better than GPT-5.2-Codex?
- Claude Fable 5 leads on quality: 70.5 vs 60.8. The BenchLeader Index combines every independent quality benchmark; Claude Fable 5 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is Claude Fable 5 better than GPT-5.2-Codex for coding?
- Claude Fable 5 scores higher in coding (71 vs 56 on the category index, where 50 is average).
- Is Claude Fable 5 better than GPT-5.2-Codex for agentic tasks?
- Claude Fable 5 scores higher in agentic tasks (74 vs 64 on the category index, where 50 is average).
- Which is cheaper, Claude Fable 5 or GPT-5.2-Codex?
- GPT-5.2-Codex is cheaper: $4.81 against $20.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Fable 5 or GPT-5.2-Codex?
- Claude Fable 5 streams faster: 63 against 25 output tokens per second.
- Which has the larger context window?
- Claude Fable 5 accepts more context: 1M against 400k tokens.