Qwen3 235B A22B Instruct 2507 vs MiMo V2.6 Flash RL
MiMo V2.6 Flash RL leads llmrun Score 131.9 to 98.5; neither fits 24 GB.
Run MiMo V2.6 Flash RL in the cloud →
Model A · Alibaba
Qwen3 235B A22B Instruct 2507235.1B · 22.2B active · 256K context · Apache 2.0 · Jul 2025
Model B · Xiaomi
MiMo V2.6 Flash RL309B · 15B active · 1M context · MIT · Sep 2026
Choose Qwen3 235B A22B Instruct 2507 if you want…
Fits a Mac 128 GB
Q3_K_L · 122.4 GB; MiMo V2.6 Flash RL needs 93.8 GB at its smallest quant
Lighter on memory
141.6 vs 185.8 GB at Q4_K_M
Choose MiMo V2.6 Flash RL if you want…
Higher llmrun Score
131.9 vs 98.5
Stronger at coding
123.6 vs 101.6
Longer context window
1M vs 256K tokens
Newer release
Sep 2026 vs Jul 2025
On your hardware
Each model at its default quant (Q4_K_M), a smaller one when it must. Estimates, ±20%.
8 GB
RTX 4060 · 8 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
12 GB
RTX 3060 12GB · 12 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
16 GB
RTX 5060 Ti 16GB · 16 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
24 GBreference
RTX 4090 · 24 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
32 GB
RTX 5090 · 32 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
48 GB
RTX A6000 · 48 GB tier
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
Mac 32 GB
Mac Mini M4 · Macs
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
Mac 64 GB
Mac Studio M4 Max · Macs
Qwen3 235B A22B Instruct 2507
Doesn't fit
smallest quant needs 72.4 GB
MiMo V2.6 Flash RL
Doesn't fit
smallest quant needs 93.8 GB
Mac 128 GB
Mac Studio M4 Max · Macs
Qwen3 235B A22B Instruct 2507
Q3_K_L · ~21 tok/s
122.4 GB · up to 32Kbetter quant
MiMo V2.6 Flash RL
IQ2_M · ~44 tok/s
105.4 GB · up to 256K
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
Qwen3 235B A22B Instruct 2507
MiMo V2.6 Flash RL
Coding
Qwen3 235B A22B Instruct 2507
MiMo V2.6 Flash RL
Science
Qwen3 235B A22B Instruct 2507
no data
MiMo V2.6 Flash RL
Reasoning
Qwen3 235B A22B Instruct 2507
MiMo V2.6 Flash RL
no data
Qwen3 235B A22B Instruct 2507 · #63 of 106 open models · #41 of 82 that fit ≤192 GB
MiMo V2.6 Flash RL · #5 of 106 open models · #1 of 82 that fit ≤192 GB
14 more boards (other shared and single-model boards)
Qwen3 235B A22B Instruct 2507
1422.2#33/188 open
ArenaLMArena · Oct 2026
MiMo V2.6 Flash RL
1447.6#16/188 open
ArenaLMArena · Oct 2026
Only Qwen3 235B A22B Instruct 2507 (8)
- ARC-AGI-21.3%
- MMLU-Pro83.0%
- Aider Polyglot59.6%
- ARC-AGI11.0%
- WeirdML38.7%
- DTBench78.4%
- Fiction.LiveBench52.9%
- LMCA23.5%
Only MiMo V2.6 Flash RL (5)
- LMArena WebDev1637.2
- SciCode51.3%
- CritPt12.0%
- ALE-Bench1244.2
- ProofBench63.0%
VRAM by quantization
Memory to load the model and 2K context; below it, the total at 32K and 128K context. Lighter side in bold.
Q2_K
Qwen3 235B A22B Instruct 2507
100 GB85.7 GB file
32K 103 GB · 128K 113 GB
MiMo V2.6 Flash RL
132 GB126.4 GB file
32K 133 GB · 128K 138 GB
Q3_K_S
Qwen3 235B A22B Instruct 2507
103 GB101.4 GB file
32K 106 GB · 128K 116 GB
MiMo V2.6 Flash RL
not listed
Q3_K_M
Qwen3 235B A22B Instruct 2507
115 GB112.4 GB file
32K 118 GB · 128K 128 GB
MiMo V2.6 Flash RL
151 GB~150.6 GB file
32K 153 GB · 128K 157 GB
Q4_0
Qwen3 235B A22B Instruct 2507
118 GB133.1 GB file
32K 121 GB · 128K 130 GB
MiMo V2.6 Flash RL
155 GB~154.5 GB file
32K 156 GB · 128K 161 GB
Q4_K_M
Qwen3 235B A22B Instruct 2507
142 GB142.2 GB file
32K 145 GB · 128K 154 GB
MiMo V2.6 Flash RL
186 GB~185.4 GB file
32K 187 GB · 128K 192 GB
Q5_K_M
Qwen3 235B A22B Instruct 2507
168 GB166.8 GB file
32K 171 GB · 128K 180 GB
MiMo V2.6 Flash RL
221 GB~220.2 GB file
32K 222 GB · 128K 227 GB
Q6_K
Qwen3 235B A22B Instruct 2507
194 GB193.0 GB file
32K 197 GB · 128K 207 GB
MiMo V2.6 Flash RL
255 GB~254.9 GB file
32K 257 GB · 128K 262 GB
Q8_0
Qwen3 235B A22B Instruct 2507
236 GB249.9 GB file
32K 239 GB · 128K 248 GB
MiMo V2.6 Flash RL
309 GB~309.0 GB file
32K 311 GB · 128K 316 GB
15 more quants
IQ2_XS
Qwen3 235B A22B Instruct 2507
71.0 GB65.6 GB file
32K 74.0 GB · 128K 83.5 GB
MiMo V2.6 Flash RL
93.1 GB98.7 GB file
32K 94.6 GB · 128K 99.4 GB
IQ2_S
Qwen3 235B A22B Instruct 2507
74.0 GB66.1 GB file
32K 76.9 GB · 128K 86.4 GB
MiMo V2.6 Flash RL
not listed
IQ2_M
Qwen3 235B A22B Instruct 2507
79.8 GB78.6 GB file
32K 82.8 GB · 128K 92.3 GB
MiMo V2.6 Flash RL
105 GB107.8 GB file
32K 106 GB · 128K 111 GB
IQ3_XXS
Qwen3 235B A22B Instruct 2507
91.6 GB93.1 GB file
32K 94.6 GB · 128K 104 GB
MiMo V2.6 Flash RL
not listed
IQ3_XS
Qwen3 235B A22B Instruct 2507
97.5 GB96.9 GB file
32K 100 GB · 128K 110 GB
MiMo V2.6 Flash RL
not listed
IQ3_S
Qwen3 235B A22B Instruct 2507
not listed
MiMo V2.6 Flash RL
132 GB146.8 GB file
32K 133 GB · 128K 138 GB
IQ3_M
Qwen3 235B A22B Instruct 2507
106 GB107.3 GB file
32K 109 GB · 128K 119 GB
MiMo V2.6 Flash RL
139 GB148.6 GB file
32K 141 GB · 128K 146 GB
Q3_K_L
Qwen3 235B A22B Instruct 2507
121 GB111.2 GB file
32K 124 GB · 128K 133 GB
MiMo V2.6 Flash RL
not listed
IQ4_XS
Qwen3 235B A22B Instruct 2507
127 GB125.5 GB file
32K 130 GB · 128K 139 GB
MiMo V2.6 Flash RL
not listed
IQ4_NL
Qwen3 235B A22B Instruct 2507
133 GB133.1 GB file
32K 136 GB · 128K 145 GB
MiMo V2.6 Flash RL
not listed
Q4_1
Qwen3 235B A22B Instruct 2507
133 GB147.2 GB file
32K 136 GB · 128K 145 GB
MiMo V2.6 Flash RL
not listed
Q4_K_S
Qwen3 235B A22B Instruct 2507
133 GB133.7 GB file
32K 136 GB · 128K 145 GB
MiMo V2.6 Flash RL
not listed
Q4_K_L
Qwen3 235B A22B Instruct 2507
144 GB143.1 GB file
32K 147 GB · 128K 157 GB
MiMo V2.6 Flash RL
not listed
Q5_K_S
Qwen3 235B A22B Instruct 2507
162 GB161.9 GB file
32K 165 GB · 128K 175 GB
MiMo V2.6 Flash RL
not listed
BF16
Qwen3 235B A22B Instruct 2507
471 GB470.3 GB file
32K 474 GB · 128K 483 GB
MiMo V2.6 Flash RL
618 GB~618.0 GB file
32K 620 GB · 128K 625 GB
Qwen3 235B A22B Instruct 2507
Ollama · Q4_K_M
ollama run hf.co/unsloth/Qwen3-235B-A22B-Instruct-2507-GGUF:Q4_K_MMiMo V2.6 Flash RL
Ollama · Q4_K_M
ollama run hf.co/vultr/MiMo-V2.6-Flash-RL-GGUF| Spec | Qwen3 235B A22B Instruct 2507 | MiMo V2.6 Flash RL |
|---|---|---|
| Publisher | Alibaba | Xiaomi |
| Parameters | 235.1B | 309B |
| Active parameters | 22B per token | 15B per token |
| Architecture | Qwen3MoeForCausalLM | MiMoV2ForCausalLM |
| Context | 256K tokens | 1M tokens |
| Modalities | Chat, Function-calling | Chat, Function-calling |
| License | Apache 2.0 | MIT |
| Released | Jul 2025 | Sep 2026 |
Frequently Asked Questions
Which is better, Qwen3 235B A22B Instruct 2507 or MiMo V2.6 Flash RL?
MiMo V2.6 Flash RL leads llmrun Score 131.9 to 98.5; neither fits 24 GB. Choose Qwen3 235B A22B Instruct 2507 for fits a Mac 128 GB and lighter on memory; choose MiMo V2.6 Flash RL for higher llmrun Score and stronger at coding.
Which needs less VRAM, Qwen3 235B A22B Instruct 2507 or MiMo V2.6 Flash RL?
At Q4_K_M, Qwen3 235B A22B Instruct 2507 requires 141.6 GB while MiMo V2.6 Flash RL needs 185.8 GB, making Qwen3 235B A22B Instruct 2507 the easier model to run on your local setup.
Can I run Qwen3 235B A22B Instruct 2507 or MiMo V2.6 Flash RL on a 16 GB GPU?
Qwen3 235B A22B Instruct 2507 does not fit (smallest quant needs 72.4 GB); MiMo V2.6 Flash RL does not fit (smallest quant needs 93.8 GB).
Which has a longer context window, Qwen3 235B A22B Instruct 2507 or MiMo V2.6 Flash RL?
Qwen3 235B A22B Instruct 2507 supports 262,144 tokens and MiMo V2.6 Flash RL supports 1,048,576 tokens.