Model comparison
Claude Fable 5 vs Grok 4.6
Claude Fable 5 is the stronger model overall, scoring 66.8 to 56.9 on the Noometry Index. Grok 4.6 costs 6.7× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Last verified . 48 shared benchmarks.
Summary
- They share 48 benchmarks with published results for both. Claude Fable 5 scores higher in 9 categories and Grok 4.6 in 1 category; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Fable 5 leads 88.5 to 67.0.
- The biggest single-benchmark swing is FrontierMath Tier 4: 90.2% for Claude Fable 5 and 31.7% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $10 / $50 for Claude Fable 5.
- Claude Fable 5 accepts more context: 1M tokens versus 500K.
Side by side
| Claude Fable 5 | Grok 4.6 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 66.8 | 56.9 |
| Released | 2026-06-07 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $10 | $2 |
| Output $ / M tokens | $50 | $6 |
| Results tracked | 62 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Fable 5 leads
Claude Fable 5: 70.6 (#4), Grok 4.6: 58.5 (#16)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| DeepSWE | 69.9% | 67.5% |
| FrontierCode | 53.5% | 48% |
| LMArena WebDev | 1625 | 1617 |
| FrontierSWE | 47% | 25.3% |
| SciCode | 61% | 56.5% |
| WeirdML | 91.9% | 67.3% |
| LMArena Coding | 1519 | 1465 |
| ALE-Bench | 2,041 | 1,508 |
| CursorBench | — | 41.4% |
| GSO | 78.4% | — |
| MirrorCode | 63.9% | — |
Agentic & Tool Use Claude Fable 5 leads
Claude Fable 5: 54.0 (#2), Grok 4.6: 39.4 (#27)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| APEX-Agents | 63.6% | 65.3% |
| GDP.pdf | 30% | 17.2% |
| Vending-Bench 2 | 5,680 | 9,047 |
| Remote Labor Index | 16.1% | — |
| τ²-bench Banking | 39.7% | — |
| PostTrainBench | 41.8% | — |
| GBAEval | 74.5% | — |
| LMArena Search | 1230 | — |
Reasoning Claude Fable 5 leads
Claude Fable 5: 76.8 (#6), Grok 4.6: 61.4 (#20)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 89.2% | 67.1% |
| SimpleBench | 81.9% | 75.9% |
| NYT Connections (extended) | 92.7% | 80% |
| ARC-AGI-1 | 98.5% | 87.5% |
| CritPt | 28.6% | 19.7% |
| Chess Puzzles | 41% | 40% |
| EBR-Bench | 39.5% | 30.5% |
| LMArena Hard Prompts | 1508 | 1447 |
| Mystery Game Puzzles | 52% | 34% |
| DTBench | 98.4% | 97.3% |
| LMCA | 61.1% | 48.5% |
| Epoch Capabilities Index | 162.06 | 156.44 |
| Kagi LLM Benchmark | 91.4% | — |
| EnigmaEval | 39.3% | — |
| Surface Evolver Bench | 95% | — |
| Bench to the Future 3 | 0.13 | — |
Math Claude Fable 5 leads
Claude Fable 5: 88.5 (#5), Grok 4.6: 67.0 (#24)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 87% | 66% |
| FrontierMath Tier 4 | 90.2% | 31.7% |
| OTIS Mock AIME 2024-2025 | 100% | 99.2% |
| ProofBench | 95% | 51% |
| LMArena Math | 1519 | 1423 |
| FrontierMath Erdős | 0% | — |
Knowledge Grok 4.6 leads
Claude Fable 5: 62.2 (#25), Grok 4.6: 63.3 (#20)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 85.9% | 94% |
| SimpleQA Verified | 70.7% | 49.3% |
| LMArena Expert | 1534 | 1467 |
Multimodal Claude Fable 5 leads
Claude Fable 5: 45.3 (#17), Grok 4.6: 43.6 (#23)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1324 | 1263 |
| Blueprint-Bench 2 | 38.6% | 33.2% |
| Furniture Assembly | 35.8% | 40% |
| LMArena Document | 1496 | 1452 |
Multilingual Claude Fable 5 leads
Claude Fable 5: 57.3 (#9), Grok 4.6: 53.0 (#74)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1481 | 1420 |
| LMArena Chinese | 1543 | 1480 |
| LMArena French | 1505 | 1461 |
| LMArena German | 1486 | 1431 |
| LMArena Japanese | 1506 | 1376 |
| LMArena Korean | 1488 | 1397 |
| LMArena Russian | 1504 | 1422 |
| LMArena Spanish | 1498 | 1404 |
Instruction Following Claude Fable 5 leads
Claude Fable 5: 78.6 (#8), Grok 4.6: 75.4 (#63)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1502 | 1431 |
Long Context Claude Fable 5 leads
Claude Fable 5: 46.3 (#23), Grok 4.6: 44.5 (#66)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1509 | 1454 |
Writing & Preference Claude Fable 5 leads
Claude Fable 5: 75.9 (#5), Grok 4.6: 62.3 (#80)
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1491 | 1428 |
| LMArena Creative Writing | 1494 | 1428 |
| LMArena Multi-Turn | 1504 | 1425 |
| EQ-Bench Creative Writing | 1943 | — |
| EQ-Bench 4 | 1340 | — |
Frequently asked questions
Is Claude Fable 5 better than Grok 4.6?
Claude Fable 5 is the stronger model overall, scoring 66.8 to 56.9 on the Noometry Index. Grok 4.6 costs 6.7× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Which is cheaper, Claude Fable 5 or Grok 4.6?
Grok 4.6 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Fable 5 lists at $10 and $50.
Is Claude Fable 5 or Grok 4.6 better for coding?
Claude Fable 5 scores higher on coding benchmarks: 70.6 versus 58.5 in the Noometry coding category.
Which has the bigger context window?
Claude Fable 5 does, with 1M tokens against 500K.
How many benchmarks do Claude Fable 5 and Grok 4.6 share?
48 benchmarks have published results for both models. Claude Fable 5 has 62 scored results on Noometry and Grok 4.6 has 49.