MiMo-V2.6-Flash
Xiaomi · released
The lighter MiMo-V2.6 model: the same multimodal inputs and million-token context at roughly a third of the size and a fraction of the price.
Best for: Very low-cost inference · Open weights · Multimodal input
- Intelligence Index
- 37.9
- #22 of 38
- Price per 1M tokens
- $0.14 in $0.28 out
- Blended $0.18 at 3:1
- Context window
- 1M
- Up to 131K output
- Output speed
- 58 tokens/s
- $0.06 measured cost per task
Capability profile
One 0–100 score per area, placing this model between the weakest and strongest tracked model across that area's benchmarks.
- Overall60 · #20 of 38
Based on 3 of 3 benchmarks
- Coding54 · #18 of 39
Based on 5 of 5 benchmarks
- Agents & knowledge work64 · #18 of 40
Based on 6 of 6 benchmarks
- Reasoning & science43 · #23 of 39
Based on 2 of 2 benchmarks
- Knowledge & reliability49 · #32 of 39
Based on 2 of 2 benchmarks
- Vision51 · #25 of 31
Based on 2 of 2 benchmarks
Benchmark results
All third-party results on record. Rank is among the tracked models that have been tested on that benchmark.
Overall
#22 of 38· Artificial Analysis
- Vals Index53.2%
#13 of 26· $0.20 per test· Vals AI
- Arena · Text1452 ±8
#22 of 33· 6,053 votes· Arena
Coding
- Terminal-Bench 4.022.7%
#19 of 38· Artificial Analysis
- SciCode51.3%
#22 of 38· Artificial Analysis
- Arena · WebDev1636 ±11
#11 of 34· 3,851 votes· Arena
- Vibe Code Bench79.0%
#15 of 26· Vals AI
- Code Migration40.4%
#17 of 26· Vals AI
Agents & knowledge work
- GDPval-AA55.5
#13 of 39· Artificial Analysis
- Arena · Agent-0.6 ±1.3
#21 of 29· Arena
- Finance Agent v256.3%
#8 of 26· Vals AI
#13 of 26· Vals AI
- Legal Research Bench38.0%
#20 of 26· Vals AI
- Tax Agent Bench59.9%
#19 of 26· Vals AI
Reasoning & science
- Humanity's Last Exam35.1%
#24 of 39· Artificial Analysis
- CritPt12.0%
#22 of 39· Artificial Analysis
- ARC-AGI-2saturatingNot tested
- ARC-AGI-3Not tested
- GPQA DiamondsaturatingNot tested
Knowledge & reliability
#32 of 39· Artificial Analysis
- AA-LCR74.3%
#33 of 39· Artificial Analysis
Vision
- MMMU-Pro73.1%
#23 of 27· Artificial Analysis
- Arena · Vision1247 ±13
#14 of 19· 2,275 votes· Arena
Specifications
- Lab
- Xiaomi
- Released
- Sep 22, 2026
- Availability
- Generally available
- Context window
- 1,048,576 tokens
- Max output
- 131,072 tokens
- Accepts
- Text, Images, Audio, Video
- Knowledge cutoff
- Not published
- Weights
- Open
- Licence
- MIT
- Parameters
- 309B total · 15B active
- API model ID
mimo-v2.6-flash
Pricing
- Input / 1M tokens
- $0.14
- Cached input / 1M
- <$0.01
- Output / 1M tokens
- $0.28
- Blended (3:1)
- $0.18
- Measured cost / task
- $0.06
Worth knowing
- Release details come from third-party coverage and the model listings; check the model card before relying on the licence terms.
Official sources
Closest alternatives
The models nearest to this one on overall score.
23 benchmarks tracked. See methodology and sources.