Model comparison
Claude Fable 5 vs Phi-4
Claude Fable 5 is the stronger model overall, scoring 66.3 to 30.9 on the Noometry Index. Phi-4 costs 229× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
MicrosoftSummary
- They share 21 benchmarks with published results for both. Claude Fable 5 scores higher in 9 categories and Phi-4 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Fable 5 leads 88.4 to 19.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for Claude Fable 5 and 13.8% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $10 / $50 for Claude Fable 5.
- Claude Fable 5 accepts more context: 1M tokens versus 128K.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Claude Fable 5 | Phi-4 | |
|---|---|---|
| Provider | Anthropic | Microsoft |
| Noometry Index | 66.3 | 30.9 |
| Released | 2026-06-07 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | 1M | 128K |
| Max output | 128K | 4K |
| Input $ / M tokens | $10 | $0.07 |
| Output $ / M tokens | $50 | $0.14 |
| Results tracked | 61 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Fable 5 leads
Claude Fable 5: 67.0 (#4), Phi-4: 33.8 (#237)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Coding | 1519 | 1231 |
| DeepSWE | 69.9% | — |
| FrontierCode | 53.5% | — |
| LMArena WebDev | 1626 | — |
| FrontierSWE | 47% | — |
| SciCode | 61% | — |
| WeirdML | 91.9% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| MirrorCode | 63.9% | — |
| BigCodeBench Complete | — | 55.4% |
| ALE-Bench | 2,041 | — |
Agentic & Tool Use Claude Fable 5 leads
Claude Fable 5: 54.1 (#2), Phi-4: 22.6 (#120)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| APEX-Agents | 63.6% | — |
| Berkeley Function Calling Leaderboard | — | 28.8% |
| Remote Labor Index | 16.1% | — |
| τ²-bench Banking | 39.7% | — |
| PostTrainBench | 41.8% | — |
| BALROG | — | 11.6% |
| GBAEval | 74.5% | — |
| GDP.pdf | 30% | — |
| LMArena Search | 1230 | — |
| Vending-Bench 2 | 5,680 | — |
Reasoning Claude Fable 5 leads
Claude Fable 5: 76.8 (#4), Phi-4: 17.0 (#287)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| Chess Puzzles | 41% | 1% |
| LMArena Hard Prompts | 1508 | 1220 |
| Epoch Capabilities Index | 163.6 | 130.43 |
| ARC-AGI-2 | 89.2% | — |
| SimpleBench | 81.9% | — |
| Kagi LLM Benchmark | 91.4% | — |
| NYT Connections (extended) | 92.7% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 28.6% | — |
| EnigmaEval | 39.3% | — |
| EBR-Bench | 39.5% | — |
| LiveBench Reasoning | — | 47.8% |
| Mystery Game Puzzles | 52% | — |
| DTBench | 98.4% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 61.1% | — |
| Surface Evolver Bench | 95% | — |
| Bench to the Future 3 | 0.13 | — |
| LiveBench | — | 41.6% |
Math Claude Fable 5 leads
Claude Fable 5: 88.4 (#5), Phi-4: 19.6 (#279)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 13.8% |
| LMArena Math | 1519 | 1246 |
| FrontierMath (Tiers 1-3) | 87% | — |
| FrontierMath Tier 4 | 90.2% | — |
| ProofBench | 95% | — |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
| FrontierMath Erdős | 0% | — |
Knowledge Claude Fable 5 leads
Claude Fable 5: 62.3 (#22), Phi-4: 32.5 (#204)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| GPQA Diamond | 85.9% | 56.1% |
| LMArena Expert | 1534 | 1203 |
| SimpleQA Verified | 70.7% | — |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |
Multimodal Not comparable
Claude Fable 5: 45.6 (#14), Phi-4: —
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Vision | 1326 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 35.8% | — |
| LMArena Document | 1496 | — |
Multilingual Claude Fable 5 leads
Claude Fable 5: 57.6 (#8), Phi-4: 37.4 (#232)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Non-English | 1483 | 1197 |
| LMArena Chinese | 1546 | 1211 |
| LMArena French | 1509 | 1223 |
| LMArena German | 1486 | 1222 |
| LMArena Japanese | 1507 | 1158 |
| LMArena Korean | 1493 | 1151 |
| LMArena Russian | 1507 | 1209 |
| LMArena Spanish | 1503 | 1234 |
Instruction Following Claude Fable 5 leads
Claude Fable 5: 78.8 (#7), Phi-4: 60.6 (#246)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1503 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Claude Fable 5 leads
Claude Fable 5: 46.3 (#22), Phi-4: 36.9 (#221)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1510 | 1217 |
Writing & Preference Claude Fable 5 leads
Claude Fable 5: 76.1 (#5), Phi-4: 40.7 (#239)
| Benchmark | Claude Fable 5 | Phi-4 |
|---|---|---|
| LMArena Text | 1491 | 1217 |
| LMArena Creative Writing | 1494 | 1182 |
| LMArena Multi-Turn | 1505 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench Creative Writing | 1943 | — |
| EQ-Bench 4 | 1340 | — |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Claude Fable 5 better than Phi-4?
Claude Fable 5 is the stronger model overall, scoring 66.3 to 30.9 on the Noometry Index. Phi-4 costs 229× 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 Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Claude Fable 5 lists at $10 and $50.
Is Claude Fable 5 or Phi-4 better for coding?
Claude Fable 5 scores higher on coding benchmarks: 67.0 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
Claude Fable 5 does, with 1M tokens against 128K.
How many benchmarks do Claude Fable 5 and Phi-4 share?
21 benchmarks have published results for both models. Claude Fable 5 has 61 scored results on Noometry and Phi-4 has 37.