Claude Fable 5 vs Gemini 4 Argon: Benchmark Comparison
Claude Fable 5 has the higher score on 3 of 13 shared benchmarks; Gemini 4 Argon leads on 10.
The largest observed score gap is 28.57 pts on Terminal-Bench-Science , where Gemini 4 Argon leads.
Reported ±1 standard-error ranges overlap on 9 of 12 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.
Shared benchmark results
Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.
| Benchmark | Claude Fable 5 | Gemini 4 Argon | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 24.32% | 30.55% | 6.23 pts | Uncertainty comparison unavailable |
| Legal Research Bench | 49.52% ±3.48 | 54.81% ±3.46 | 5.29 pts | Reported ±1 SE ranges overlap |
| LegalBench | 88.56% ±0.33 | 88.30% ±0.37 | 0.26 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 38.28% ±3.50 | 44.56% ±3.59 | 6.28 pts | Reported ±1 SE ranges overlap |
| MedCode | 56.07% ±2.20 | 58.80% ±2.10 | 2.73 pts | Reported ±1 SE ranges overlap |
| MedScribe | 88.52% ±1.95 | 87.43% ±1.96 | 1.09 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 95.00% ±2.19 | 99.00% ±1.00 | 4.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 15.71% ±4.38 | 44.29% ±5.98 | 28.57 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 51.89% ±3.40 | 53.65% ±3.38 | 1.76 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 2.00% ±0.99 | 2.50% ±1.11 | 0.50 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 41.41% ±1.34 | 57.58% ±2.31 | 16.16 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.35% ±2.10 | 91.91% ±1.90 | 1.55 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 70.43% ±1.19 | 69.76% ±0.00 | 0.68 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Claude Fable 5 average | Gemini 4 Argon average |
|---|---|---|
| Legal | 69.04% | 71.55% |
| Finance | 38.28% | 44.56% |
| Healthcare | 72.30% | 73.11% |
| Math | 95.00% | 99.00% |
| Science | 15.71% | 44.29% |
| Education | 51.89% | 53.65% |
| Coding | 44.59% | 50.66% |
| Social Mobility | 70.43% | 69.76% |
Cost and latency
Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.
| Benchmark | Claude Fable 5 cost | Gemini 4 Argon cost | Claude Fable 5 latency | Gemini 4 Argon latency |
|---|---|---|---|---|
| Vals RSI Index | $491.22 | $179.80 | 90h00m | 90h00m |
| Legal Research Bench | $9.79 | $6.45 | 22m18s | 31m21s |
| LegalBench | N/A | N/A | 8.96s | 12.62s |
| Tax Agent Bench | $6.70 | $3.47 | 14m27s | 16m26s |
| MedCode | N/A | N/A | 91.44s | 79.53s |
| MedScribe | N/A | N/A | 119.47s | 87.44s |
| ProofBench v1.1 | $5.45 | $0.78 | 13m43s | 6m20s |
| Terminal-Bench-Science | $51.99 | $44.90 | 1h59m | 1h46m |
| SAGE | N/A | N/A | 116.92s | 63.07s |
| ProgramBench | $75.68 | $17.66 | 2h37m | 1h11m |
| Terminal-Bench 4.0 | $30.34 | $17.64 | 1h08m | 53m35s |
| Vibe Code Bench v1.1 | $41.71 | $8.11 | 1h02m | 37m28s |
| Public Benefits Bench v1.1 | $4.59 | $2.79 | 21m17s | 26m07s |
Results available only for Claude Fable 5
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SWE-bench
Results available only for Gemini 4 Argon
- Vals Finance Index
- Vals Index
- Vals Legal Index
- Vals Science Index
- HLAB-Vals-1.1
- EMB v1.1
- Finance Agent v2.1
- BioMysteryBench
- MysteryMechanism
- Robo-bench
- Code Migration v1.1
- IOI
- SWE-bench Multimodal
- Vibe Code Bench 1-100
- CyberBench v1.1
- SRE Bench
- CUA-bench
- Time Horizon Index: KSP