SShogo Bench

MiMo-V2.6-Flash

MiMo Flash model for multimodal coding agents and long-context automation

Shogo Agent Index

Unranked

Not ranked yet: 0 agentic benchmarks measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.

Agentic coding
56
estimated
Long-horizon autonomy
60
estimated
Knowledge-work agents
40
estimated
Computer use
39
estimated
Security & context learning
50
estimated

Price

Input
$0.14/M
Output
$0.28/M
Cache read
$0.003/M
3:1 blend
$0.18/M
Agent-effective
$0.040/M

Cheapest host for agents: Darkbloom at $0.030/M.

Agent-effective = 20 input tokens per output token, 85% cache hits. Why.

Agentic benchmarks

BenchmarkScoreNormalisedIndex weightHarness / source
SWE-bench Verifiedest.8%predicted 79
Terminal-Bench 2.0est.8%predicted 77
DeepSWEest.6%predicted 53
FrontierSWEest.5%predicted 19
CursorBenchest.4%predicted 39
GSO-Benchest.4%predicted 34
METR Time Horizonsest.10%predicted 77
Vending-Bench 2est.6%predicted 55
PostTrainBenchest.4%predicted 26
APEX-Agentsest.8%predicted 49
GDPvalest.7%predicted 50
Remote Labor Indexest.5%predicted 5
DeepResearch Benchest.5%predicted 49
OSWorld 2.0est.6%predicted 12
OSWorldest.4%predicted 80
Cybenchest.5%predicted 78
CL-benchest.5%predicted 23

Other benchmarks

BenchmarkCategoryScoreSource
CritPtReasoning12.0%source
ProofBenchMath63.0%source
SciCodeCoding51.3%source

API hosts

Live prices and uptime for every host serving MiMo-V2.6-Flash via OpenRouter.

HostInputOutputCache readAgent $/MUptime 24hQuant
Darkbloom$0.10$0.28$0.003$0.03097.9%fp4
GMICloud$0.11$0.22$0.002$0.03063.9%bf16
Cohere$0.14$0.28$0.003$0.04099.9%fp8
DeepInfra$0.14$0.28$0.003$0.04099.0%fp8
Venice$0.17$0.35$0.004$0.04098.3%fp8
Xiaomi$0.14$0.28$0.003$0.040100.0%fp8
Makora$0.20$0.40$0.015$0.06097.0%fp8