Comparison set Choose two to four models, then compare their evidence.
Selected models 2 of 4
Claude Opus 5.5 MiMo-V2.6-Pro Decision evidence Exact source values, missing states, and pair-derived deltas stay separate.
Exact capability values Evidence key:
Capability scores per model Capability Claude Opus 5.5 MiMo-V2.6-Pro Agentic 65.3 77.8 Coding 87.5 63.3 Knowledge 85.4 58.4 Reasoning 86.1 82.5 Multimodal & Grounded 92.2 Unavailable Instruction Following Unavailable Unavailable Math 97.3 Unavailable Overall 86.68 73.6
Runtime Time to first token and throughput, exactly as the runtime source published them.
Time to first token: latency to the first answer chunk (Artificial Analysis, via BenchLM). Reasoning models include thinking time, so values can run to tens or hundreds of seconds.
Economics Source rates plus a labelled 75% input / 25% output blended cost.
Runtime and price matrix Runtime and price values per model Metric Claude Opus 5.5
MiMo-V2.6-Pro
Input / 1M $4.00 $0.44 Output / 1M $20.00 $0.87 Cache read / 1M $0.20 $0.0036 Cache write / 1M $5.00 Unavailable Blended / 1M (75% input / 25% output) $8.00 $0.54 Cost per successful task (LiveBench) Unavailable Unavailable Time to first token 573.15 s 43.15 s Throughput 95 tok/s 50 tok/s Context window 1M tokens 1M tokens Release date 2026-09-22 2026-09-22 Access Proprietary Open Weight
Decision deltas Each row comes from the pair comparison source. A null basis or average stays unavailable rather than being inferred from the capability table.
No pair comparison was published for this selection. The comparison source only publishes deltas for pairs it has evaluated. The capability and price tables above remain source-exact.
ModelScale