Model comparison
Claude Fable 5 vs Gemini 3.8 Flash
Claude Fable 5 is the stronger model overall, scoring 66.8 to 61.8 on the Noometry Index. Gemini 3.8 Flash costs 13× 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.
GoogleSummary
- They share 48 benchmarks with published results for both. Claude Fable 5 scores higher in 7 categories and Gemini 3.8 Flash in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Fable 5 leads 88.5 to 65.3.
- The biggest single-benchmark swing is FrontierMath Tier 4: 90.2% for Claude Fable 5 and 22% for Gemini 3.8 Flash.
- Gemini 3.8 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $10 / $50 for Claude Fable 5.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Fable 5 | Gemini 3.8 Flash | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 66.8 | 61.8 |
| Released | 2026-06-07 | 2026-09-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 66K |
| Input $ / M tokens | $10 | $0.75 |
| Output $ / M tokens | $50 | $3.75 |
| Results tracked | 62 | 50 |
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), Gemini 3.8 Flash: 59.2 (#15)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| DeepSWE | 69.9% | 73.8% |
| FrontierCode | 53.5% | 41.2% |
| LMArena WebDev | 1625 | 1584 |
| FrontierSWE | 47% | 19.6% |
| SciCode | 61% | 56.6% |
| WeirdML | 91.9% | 84.8% |
| LMArena Coding | 1519 | 1510 |
| ALE-Bench | 2,041 | 1,270 |
| CursorBench | — | 39.6% |
| GSO | 78.4% | — |
| MirrorCode | 63.9% | — |
Agentic & Tool Use Claude Fable 5 leads
Claude Fable 5: 54.0 (#2), Gemini 3.8 Flash: 41.8 (#21)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| APEX-Agents | 63.6% | 64.3% |
| Remote Labor Index | 16.1% | 5.8% |
| GDP.pdf | 30% | 23.4% |
| Vending-Bench 2 | 5,680 | 5,094 |
| τ²-bench Banking | 39.7% | — |
| PostTrainBench | 41.8% | — |
| GBAEval | 74.5% | — |
| LMArena Search | 1230 | — |
Reasoning Too close to call
Claude Fable 5: 76.8 (#6), Gemini 3.8 Flash: 76.9 (#5)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| ARC-AGI-2 | 89.2% | 89.2% |
| NYT Connections (extended) | 92.7% | 97.4% |
| ARC-AGI-1 | 98.5% | 98.5% |
| CritPt | 28.6% | 18.3% |
| Chess Puzzles | 41% | 61% |
| LMArena Hard Prompts | 1508 | 1508 |
| Mystery Game Puzzles | 52% | 47% |
| DTBench | 98.4% | 95.7% |
| LMCA | 61.1% | 52.9% |
| Surface Evolver Bench | 95% | 76.9% |
| Epoch Capabilities Index | 162.06 | 156.71 |
| SimpleBench | 81.9% | — |
| Kagi LLM Benchmark | 91.4% | — |
| EnigmaEval | 39.3% | — |
| EBR-Bench | 39.5% | — |
| Bench to the Future 3 | 0.13 | — |
Math Claude Fable 5 leads
Claude Fable 5: 88.5 (#5), Gemini 3.8 Flash: 65.3 (#28)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 87% | 68.4% |
| FrontierMath Tier 4 | 90.2% | 22% |
| OTIS Mock AIME 2024-2025 | 100% | 98.9% |
| ProofBench | 95% | 48% |
| LMArena Math | 1519 | 1528 |
| FrontierMath Erdős | 0% | — |
Knowledge Gemini 3.8 Flash leads
Claude Fable 5: 62.2 (#25), Gemini 3.8 Flash: 74.8 (#2)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| GPQA Diamond | 85.9% | 95.4% |
| SimpleQA Verified | 70.7% | 69.7% |
| LMArena Expert | 1534 | 1524 |
| Humanity's Last Exam | — | 44.5% |
Multimodal Claude Fable 5 leads
Claude Fable 5: 45.3 (#17), Gemini 3.8 Flash: 40.7 (#45)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Vision | 1324 | 1314 |
| Blueprint-Bench 2 | 38.6% | 38.6% |
| Furniture Assembly | 35.8% | 31.7% |
| LMArena Document | 1496 | — |
Multilingual Too close to call
Claude Fable 5: 57.3 (#9), Gemini 3.8 Flash: 58.0 (#5)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Non-English | 1481 | 1491 |
| LMArena Chinese | 1543 | 1554 |
| LMArena French | 1505 | 1498 |
| LMArena German | 1486 | 1493 |
| LMArena Japanese | 1506 | 1502 |
| LMArena Korean | 1488 | 1459 |
| LMArena Russian | 1504 | 1515 |
| LMArena Spanish | 1498 | 1485 |
Instruction Following Too close to call
Claude Fable 5: 78.6 (#8), Gemini 3.8 Flash: 78.0 (#13)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Instruction Following | 1502 | 1490 |
Long Context Too close to call
Claude Fable 5: 46.3 (#23), Gemini 3.8 Flash: 46.3 (#24)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Longer Query | 1509 | 1508 |
Writing & Preference Claude Fable 5 leads
Claude Fable 5: 75.9 (#5), Gemini 3.8 Flash: 72.2 (#15)
| Benchmark | Claude Fable 5 | Gemini 3.8 Flash |
|---|---|---|
| LMArena Text | 1491 | 1499 |
| LMArena Creative Writing | 1494 | 1492 |
| EQ-Bench Creative Writing | 1943 | 1748 |
| LMArena Multi-Turn | 1504 | 1501 |
| EQ-Bench 4 | 1340 | — |
Frequently asked questions
Is Claude Fable 5 better than Gemini 3.8 Flash?
Claude Fable 5 is the stronger model overall, scoring 66.8 to 61.8 on the Noometry Index. Gemini 3.8 Flash costs 13× 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 Gemini 3.8 Flash?
Gemini 3.8 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; Claude Fable 5 lists at $10 and $50.
Is Claude Fable 5 or Gemini 3.8 Flash better for coding?
Claude Fable 5 scores higher on coding benchmarks: 70.6 versus 59.2 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.8 Flash does, with 1.05M tokens against 1M.
How many benchmarks do Claude Fable 5 and Gemini 3.8 Flash share?
48 benchmarks have published results for both models. Claude Fable 5 has 62 scored results on Noometry and Gemini 3.8 Flash has 50.