Claude Haiku 5.5 vs MiMo V2.6 Flash: Benchmark Comparison
Claude Haiku 5.5 has the higher score on 20 of 24 shared benchmarks; MiMo V2.6 Flash leads on 4.
The largest observed score gap is 26.00 pts on ProofBench v1.1 , where Claude Haiku 5.5 leads.
Reported ±1 standard-error ranges overlap on 12 of 24 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 Haiku 5.5 | MiMo V2.6 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 57.45% ±1.80 | 57.15% ±1.39 | 0.30 pts | Reported ±1 SE ranges overlap |
| Vals Index | 58.71% ±1.28 | 54.83% ±1.11 | 3.88 pts | Reported ±1 SE ranges do not overlap |
| Vals Legal Index | 70.27% ±1.83 | 68.99% ±1.77 | 1.28 pts | Reported ±1 SE ranges overlap |
| Vals Science Index | 38.50% ±1.67 | 31.72% ±1.27 | 6.78 pts | Reported ±1 SE ranges do not overlap |
| HLAB-Vals-1.1 | 4.17% ±1.52 | 13.33% ±2.82 | 9.17 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 43.27% ±3.44 | 37.98% ±3.37 | 5.29 pts | Reported ±1 SE ranges overlap |
| EMB v1.1 | 65.48% ±3.69 | 68.88% ±2.27 | 3.41 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 65.37% ±2.06 | 63.32% ±2.08 | 2.06 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 31.92% ±3.35 | 28.51% ±3.14 | 3.41 pts | Reported ±1 SE ranges overlap |
| MedCode | 51.54% ±2.17 | 41.06% ±2.03 | 10.48 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 89.54% ±1.92 | 85.28% ±1.99 | 4.27 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 89.00% ±3.15 | 63.00% ±4.85 | 26.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 70.37% ±1.33 | 69.26% ±0.98 | 1.11 pts | Reported ±1 SE ranges overlap |
| MysteryMechanism | 35.13% ±3.21 | 21.62% ±2.77 | 13.51 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 10.00% ±3.61 | 4.29% ±2.44 | 5.71 pts | Reported ±1 SE ranges overlap |
| SAGE | 45.78% ±3.42 | 43.53% ±3.43 | 2.25 pts | Reported ±1 SE ranges overlap |
| Code Migration v1.1 | 52.36% ±3.49 | 39.55% ±3.54 | 12.81 pts | Reported ±1 SE ranges do not overlap |
| IOI | 47.28% ±10.72 | 47.67% ±4.79 | 0.39 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 2.50% ±1.11 | 0.50% ±0.50 | 2.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 36.36% ±2.31 | 24.24% ±1.51 | 12.12 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.44% ±1.73 | 78.96% ±4.04 | 11.48 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 75.48% ±5.21 | 75.36% ±5.42 | 0.12 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 5.72% ±1.44 | 4.20% ±1.24 | 1.53 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 63.19% ±1.25 | 67.59% ±1.22 | 4.40 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Haiku 5.5 average | MiMo V2.6 Flash average |
|---|---|---|
| Legal | 23.72% | 25.66% |
| Finance | 54.26% | 53.57% |
| Healthcare | 70.54% | 63.17% |
| Math | 89.00% | 63.00% |
| Science | 38.50% | 31.72% |
| Education | 45.78% | 43.53% |
| Coding | 45.79% | 38.18% |
| Cyber | 40.60% | 39.78% |
| Social Mobility | 63.19% | 67.59% |
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 Haiku 5.5 cost | MiMo V2.6 Flash cost | Claude Haiku 5.5 latency | MiMo V2.6 Flash latency |
|---|---|---|---|---|
| Vals Finance Index | $1.39 | $0.07 | 33m31s | 19m26s |
| Vals Index | $3.17 | $0.22 | 49m48s | 1h07m |
| Vals Legal Index | $2.35 | $0.07 | 58m22s | 18m20s |
| Vals Science Index | $1.86 | $0.12 | 39m58s | 1h41m |
| HLAB-Vals-1.1 | $4.26 | $0.09 | 43m41s | 22m23s |
| Legal Research Bench | $1.63 | $0.07 | 1h22m | 15m20s |
| EMB v1.1 | $1.36 | $0.11 | 19m56s | 25m22s |
| Finance Agent v2.1 | $1.64 | $0.07 | 31m05s | 15m36s |
| Tax Agent Bench | $1.16 | $0.05 | 49m34s | 17m19s |
| MedCode | N/A | N/A | 3m46s | 93.37s |
| MedScribe | N/A | N/A | 5m23s | 74.48s |
| ProofBench v1.1 | $0.17 | $0.12 | 6m58s | 1h03m |
| BioMysteryBench | $0.51 | $0.05 | 12m32s | 28m28s |
| MysteryMechanism | $0.54 | $0.06 | 13m25s | 35m08s |
| Terminal-Bench-Science | $4.52 | $0.24 | 1h34m | 3h59m |
| SAGE | N/A | N/A | 2m23s | 2m16s |
| Code Migration v1.1 | $6.20 | $0.53 | 1h26m | 4h10m |
| IOI | $0.83 | $0.22 | 12m34s | 1h49m |
| ProgramBench | $6.83 | $0.47 | 1h13m | 3h22m |
| Terminal-Bench 4.0 | $3.12 | $0.22 | 51m56s | 2h25m |
| Vibe Code Bench v1.1 | $6.07 | $0.56 | 34m36s | 50m13s |
| CyberBench v1.1 | $1.25 | $0.05 | 32m39s | 28m22s |
| SRE Bench | $3.14 | $0.40 | 57m56s | 3h22m |
| Public Benefits Bench v1.1 | $0.53 | $0.03 | 36m32s | 20m56s |
Results available only for Claude Haiku 5.5
- ClaimBench
- SWE-bench Multimodal
- Vibe Code Bench 1-100
Results available only for MiMo V2.6 Flash
None.