Mixtral 8x7B Instruct v0.1
Dec 2023
Mixtral 8x7B Instruct v0.1Compare inon
Benchmarks of Mixtral 8x7B Instruct v0.118
BenchmarkByScoreRank
Math1
Writing15
Legacy · saturated or no longer reported2
No benchmark matches
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperGLM 5.3 FlashunstatedOpenwins 83%-56%$0.15 / $0.50Memory (4-bit): 28 GB → 193 GB+5
BetterGPT-5 Minihighwins 72%+27%$0.053
OpenGemma 4 31Bunstatedwins 82%-70%$0.10 / $0.34
CheaperHy3unstatedOpenwins 82%-57%$0.13 / $0.53Memory (4-bit): 28 GB → 179 GBCheaperGPT-6 Lunamaxwins 81%-63%$0.068Function calling only with none · Chat Completions+3ModelScoreCostPer taskWhat changes
GLM 5.3 FlashunstatedOpenwins 83%-56%$0.15 / $0.50Memory (4-bit): 28 GB → 193 GB+5Hy3unstatedOpenwins 82%-57%$0.13 / $0.53Memory (4-bit): 28 GB → 179 GBGemma 4 31BunstatedOpenwins 82%-70%$0.10 / $0.34GPT-6 Lunamaxwins 81%-63%$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 81%-17%$0.085ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 72%+27%$0.053GPT-6 Lunamaxwins 81%-63%$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 81%-17%$0.085GPT-5.6 Terraxhighwins 83%+733%$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 85%+641%$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Gemma 4 31BunstatedOpenwins 82%-70%$0.10 / $0.34Gemma 4 26B A4BunstatedOpenwins 80%-51%$0.15 / $0.60
Qwen3.8 27BunstatedOpenwins 77%+77%$0.42 / $2.55max_tokens range · QwenCloud+4
Muse Glimmer 30BunstatedOpenwins 76%-7%$0.30 / $1.10Qwen 3.5 27BunstatedOpenwins 76%+53%$0.30 / $2.40ModelScoreCostPer taskWhat changes
Mistral Medium 3.5unstatedOpenwins 80%—$1.50 / $7.50Memory (4-bit): 28 GB → 77 GBMistral Large 3unstatedOpenwins 78%—$0.50 / $1.50Memory (4-bit): not verifiedMistral Large 4unstatedwins 77%—$1.36 / $4.18Infinite whitespace stream+4Mistral Largeunstatedwins 66%—$2 / $6Mistral Small 3.1 24BunstatedOpenwins 65%—$0.35 / $0.56Mistral Small 3unstatedOpenwins 60%—$0.05 / $0.08Context: 33K → 32KMixtral 8x22B InstructunstatedOpenwins 54%—$2 / $6Memory (4-bit): 28 GB → 84 GBMistral 7B Instruct v0.2Openwins 42%—$0.055 / $0.055No strict JSONAll Mistral models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✗ 16 GB✗ 24 GB GPU✓ 48 GB GPU · compressed✓ 80 GB GPU · 8-bit
LicenceApache-2.0free for commercial use
Benchmarks
all categoriesChat & Knowledge
Arena Text · overall1,197#184
Arena Text · Chinese1,055.4#175
Arena Text · English1,167.5#178
Arena Text · French1,166.2#139
Arena Text · German1,114#144
Arena Text · instruction following1,109.2#177
Arena Text · Japanese931.2#136
Arena Text · Korean967.8#140
Arena Text · non-English1,076.9#177
Arena Text · Polish1,072.1#133
Arena Text · Russian1,090.4#178
Arena Text · Spanish1,110.8#150
MATH Level 59.3#32
Arena Text · overall · full cohort1,131.5#177
GPQA Diamond · Epoch30.6#120
MMLU-Pro · official board43.3#22
Writing & Language
Arena Text · creative writing1,159#183
Arena Text · creative writing · full cohort1,108.6#176
FAQ
How much does Mixtral 8x7B Instruct v0.1 cost?›
$0.54 per 1M input tokens and $0.54 per 1M output tokens (cheapest host: DeepInfra).
Is Mixtral 8x7B Instruct v0.1 free to use?›
There is no free API; the cheapest paid price is $0.54 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Mixtral 8x7B Instruct v0.1, and what is it best at?›
#184 of 196 in Chat & Knowledge, #183 of 195 in Writing & Language.
Is Mixtral 8x7B Instruct v0.1 better than Llama 3.1 8B Instruct?›
On Arena Text · creative writing, Mixtral 8x7B Instruct v0.1 scores 1,159 vs 1,177 — lower.
How can I use Mixtral 8x7B Instruct v0.1?›
Through the API of DeepInfra.
What is the context window of Mixtral 8x7B Instruct v0.1?›
33K tokens.
When was Mixtral 8x7B Instruct v0.1 released? Is there a newer version?›
Released 2023-12-10. It is the current version in its line.
Is Mixtral 8x7B Instruct v0.1 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Mixtral 8x7B Instruct v0.1 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Mixtral 8x7B Instruct v0.1 read images?›
No, it accepts text only.
What is Mixtral 8x7B Instruct v0.1?›
Mixtral 8x7B Instruct v0.1 is an open-weights language model from Mistral, released Dec 2023.
What are the best alternatives to Mixtral 8x7B Instruct v0.1?›
GLM 5.3 Flash (cheaper), GPT-5 Mini (better), Gemma 4 31B (open weights).