Mixtral 8x7B Instruct v0.1 vs GPT OSS 20B
GPT OSS 20B leads llmrun Score 96.4 to 65.8; both fit 24 GB (Mixtral 8x7B Instruct v0.1 at IQ3_M, GPT OSS 20B at Q4_K_M); GPT OSS 20B is 2.1× faster on RTX 4090 (~190 vs ~92 tok/s).
Model A · Mistral AI
Mixtral 8x7B Instruct v0.146.7B · 12.9B active · 32K context · Apache 2.0 · Dec 2023
Model B · OpenAI
GPT OSS 20B20.9B · 4.2B active · 128K context · Apache 2.0 · Aug 2025
Choose GPT OSS 20B if you want…
Higher llmrun Score
96.4 vs 65.8
Stronger at reasoning
91.3 vs 71.0
Ahead on the benchmarks
3 of 3 shared boards
Fits a 12 GB GPU
IQ4_XS · 11.8 GB; Mixtral 8x7B Instruct v0.1 needs 15.3 GB at its smallest quant
2.1× faster on RTX 4090
~190 vs ~92 tok/s
Lighter on memory
12.9 vs 28.6 GB at Q4_K_M
Mixtral 8x7B Instruct v0.1: no clear edge over GPT OSS 20B in the data we have.
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
Mixtral 8x7B Instruct v0.1
Doesn't fit
smallest quant needs 15.3 GB
GPT OSS 20B
IQ2_M · ~103 tok/s
7.6 GB · up to 32Kfits
12 GB
RTX 3060 12GB · 12 GB tier
Mixtral 8x7B Instruct v0.1
Doesn't fit
smallest quant needs 15.3 GB
GPT OSS 20B
IQ4_XS · ~89 tok/s
11.8 GB · up to 16Kfits
16 GB
RTX 5060 Ti 16GB · 16 GB tier
Mixtral 8x7B Instruct v0.1
IQ2_XXS · ~69 tok/s
15.3 GB · up to 16K
GPT OSS 20B
Q4_K_M · ~98 tok/s
13.1 GB · up to 128Kmore headroom
24 GBreference
RTX 4090 · 24 GB tier
Mixtral 8x7B Instruct v0.1
IQ3_M · ~92 tok/s
23.5 GB · up to 16K
GPT OSS 20B
Q4_K_M · ~190 tok/s
13.1 GB · up to 128Kmore headroom
32 GB
RTX 5090 · 32 GB tier
Mixtral 8x7B Instruct v0.1
Q4_K_M · ~117 tok/s
30.5 GB · up to 16K
GPT OSS 20B
Q4_K_M · ~282 tok/s
13.1 GB · up to 128Kfaster
48 GB
RTX A6000 · 48 GB tier
Mixtral 8x7B Instruct v0.1
Q4_K_M · ~57 tok/s
30.5 GB · up to 32K
GPT OSS 20B
Q4_K_M · ~154 tok/s
13.1 GB · up to 128Kfaster
Mac 32 GB
Mac Mini M4 · Macs
Mixtral 8x7B Instruct v0.1
IQ4_XS · ~11 tok/s
27.6 GB · up to 16K
GPT OSS 20B
Q4_K_M · ~28 tok/s
13.1 GB · up to 128Kmore headroom
Mac 64 GB
Mac Studio M4 Max · Macs
Mixtral 8x7B Instruct v0.1
Q4_K_M · ~37 tok/s
30.5 GB · up to 32K
GPT OSS 20B
Q4_K_M · ~86 tok/s
13.1 GB · up to 128Kfaster
Mac 128 GB
Mac Studio M4 Max · Macs
Mixtral 8x7B Instruct v0.1
Q4_K_M · ~37 tok/s
30.5 GB · up to 32K
GPT OSS 20B
Q4_K_M · ~86 tok/s
13.1 GB · up to 128Kfaster
Reference GPU: RTX 4090. Tinted cell = the better pick on that machine (fits, more headroom, or at least 10% faster).
Overall
Mixtral 8x7B Instruct v0.1
GPT OSS 20B
Coding
Mixtral 8x7B Instruct v0.1
no data
GPT OSS 20B
Science
Mixtral 8x7B Instruct v0.1
no data
GPT OSS 20B
Reasoning
Mixtral 8x7B Instruct v0.1
GPT OSS 20B
Mixtral 8x7B Instruct v0.1 · #100 of 106 open models · #31 of 36 that fit 24 GB
GPT OSS 20B · #69 of 106 open models · #16 of 36 that fit 24 GB
Benchmarks
3 shared boards: Mixtral 8x7B Instruct v0.1 leads 0, GPT OSS 20B leads 3 · Benchmarks imported Nov 2026
Mixtral 8x7B Instruct v0.1
30.6%#83/100 open
IndependentEpoch AI · Dec 2023
GPT OSS 20B
60.8%#44/100 open
IndependentEpoch AI · Aug 2025
Mixtral 8x7B Instruct v0.1
43.3%#93/146 open
IndependentTIGER-Lab
GPT OSS 20B
73.6%#33/146 open
IndependentTIGER-Lab
Mixtral 8x7B Instruct v0.1
49.6%#62/70 open
Lab-reportedEpoch AI · Dec 2023
GPT OSS 20B
68.0%#36/70 open
Lab-reportedEpoch AI · Aug 2025
10 more boards (other shared and single-model boards)
Mixtral 8x7B Instruct v0.1
1197.2#148/188 open
ArenaLMArena · Oct 2026
GPT OSS 20B
1317.3#98/188 open
ArenaLMArena · Oct 2026
Only Mixtral 8x7B Instruct v0.1 (2)
- MATH Level 510.0%
- ForecastBench58.2
Only GPT OSS 20B (7)
- AIME 2024/202565.3%
- Terminal-Bench3.4%
- SciCode38.9%
- WeirdML40.9%
- CritPt1.4%
- LMCA14.5%
- ALE-Bench566.0
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
Mixtral 8x7B Instruct v0.1
20.4 GB15.6 GB file
32K 24.5 GB · 128K —
GPT OSS 20B
9.3 GB11.5 GB file
32K 9.8 GB · 128K 11.5 GB
Q3_K_S
Mixtral 8x7B Instruct v0.1
21.0 GB20.4 GB file
32K 25.0 GB · 128K —
GPT OSS 20B
9.5 GB11.5 GB file
32K 10.0 GB · 128K 11.7 GB
Q3_K_M
Mixtral 8x7B Instruct v0.1
23.3 GB20.4 GB file
32K 27.4 GB · 128K —
GPT OSS 20B
10.6 GB11.5 GB file
32K 11.1 GB · 128K 12.8 GB
Q4_0
Mixtral 8x7B Instruct v0.1
23.9 GB26.4 GB file
32K 28.0 GB · 128K —
GPT OSS 20B
10.8 GB11.5 GB file
32K 11.3 GB · 128K 13.0 GB
Q4_K_M
Mixtral 8x7B Instruct v0.1
28.6 GB26.4 GB file
32K 32.6 GB · 128K —
GPT OSS 20B
12.9 GB11.6 GB file
32K 13.4 GB · 128K 15.1 GB
Q5_K_M
Mixtral 8x7B Instruct v0.1
33.8 GB32.2 GB file
32K 37.9 GB · 128K —
GPT OSS 20B
15.3 GB11.7 GB file
32K 15.8 GB · 128K 17.5 GB
Q6_K
Mixtral 8x7B Instruct v0.1
39.1 GB38.4 GB file
32K 43.1 GB · 128K —
GPT OSS 20B
17.6 GB12.0 GB file
32K 18.1 GB · 128K 19.8 GB
Q8_0
Mixtral 8x7B Instruct v0.1
47.3 GB49.6 GB file
32K 51.3 GB · 128K —
GPT OSS 20B
21.3 GB~20.9 GB file
32K 21.8 GB · 128K 23.5 GB
20 more quants
IQ2_XXS
Mixtral 8x7B Instruct v0.1
13.4 GB12.6 GB file
32K 17.4 GB · 128K —
GPT OSS 20B
6.1 GB~5.8 GB file
32K 6.6 GB · 128K 8.3 GB
IQ2_XS
Mixtral 8x7B Instruct v0.1
14.6 GB13.9 GB file
32K 18.6 GB · 128K —
GPT OSS 20B
6.7 GB~6.3 GB file
32K 7.2 GB · 128K 8.9 GB
IQ2_S
Mixtral 8x7B Instruct v0.1
15.2 GB14.1 GB file
32K 19.2 GB · 128K —
GPT OSS 20B
6.9 GB~6.5 GB file
32K 7.4 GB · 128K 9.1 GB
IQ2_M
Mixtral 8x7B Instruct v0.1
16.3 GB15.5 GB file
32K 20.4 GB · 128K —
GPT OSS 20B
7.4 GB~7.1 GB file
32K 7.9 GB · 128K 9.6 GB
IQ3_XXS
Mixtral 8x7B Instruct v0.1
18.7 GB18.2 GB file
32K 22.7 GB · 128K —
GPT OSS 20B
8.5 GB11.6 GB file
32K 9.0 GB · 128K 10.7 GB
Q2_K_S
Mixtral 8x7B Instruct v0.1
19.3 GB16.0 GB file
32K 23.3 GB · 128K —
GPT OSS 20B
8.7 GB12.1 GB file
32K 9.2 GB · 128K 10.9 GB
IQ3_XS
Mixtral 8x7B Instruct v0.1
19.8 GB19.4 GB file
32K 23.9 GB · 128K —
GPT OSS 20B
9.0 GB11.6 GB file
32K 9.5 GB · 128K 11.2 GB
IQ3_S
Mixtral 8x7B Instruct v0.1
20.4 GB20.4 GB file
32K 24.5 GB · 128K —
GPT OSS 20B
9.3 GB12.1 GB file
32K 9.8 GB · 128K 11.5 GB
IQ3_M
Mixtral 8x7B Instruct v0.1
21.6 GB21.4 GB file
32K 25.6 GB · 128K —
GPT OSS 20B
9.8 GB11.6 GB file
32K 10.3 GB · 128K 12.0 GB
Q3_K_L
Mixtral 8x7B Instruct v0.1
24.5 GB24.2 GB file
32K 28.5 GB · 128K —
GPT OSS 20B
11.1 GB11.5 GB file
32K 11.6 GB · 128K 13.3 GB
IQ4_XS
Mixtral 8x7B Instruct v0.1
25.7 GB25.4 GB file
32K 29.7 GB · 128K —
GPT OSS 20B
11.6 GB11.6 GB file
32K 12.1 GB · 128K 13.8 GB
IQ4_NL
Mixtral 8x7B Instruct v0.1
26.8 GB26.5 GB file
32K 30.9 GB · 128K —
GPT OSS 20B
12.1 GB11.6 GB file
32K 12.6 GB · 128K 14.3 GB
Q4_1
Mixtral 8x7B Instruct v0.1
26.8 GB29.3 GB file
32K 30.9 GB · 128K —
GPT OSS 20B
12.1 GB11.6 GB file
32K 12.6 GB · 128K 14.3 GB
Q4_K_S
Mixtral 8x7B Instruct v0.1
26.8 GB26.7 GB file
32K 30.9 GB · 128K —
GPT OSS 20B
12.1 GB11.6 GB file
32K 12.6 GB · 128K 14.3 GB
Q4_K_L
Mixtral 8x7B Instruct v0.1
not listed
GPT OSS 20B
13.2 GB11.9 GB file
32K 13.7 GB · 128K 15.4 GB
Q5_0
Mixtral 8x7B Instruct v0.1
29.8 GB32.2 GB file
32K 33.8 GB · 128K —
GPT OSS 20B
not listed
Q5_1
Mixtral 8x7B Instruct v0.1
32.7 GB35.1 GB file
32K 36.7 GB · 128K —
GPT OSS 20B
not listed
Q5_K_S
Mixtral 8x7B Instruct v0.1
32.7 GB32.2 GB file
32K 36.7 GB · 128K —
GPT OSS 20B
14.8 GB11.7 GB file
32K 15.3 GB · 128K 17.0 GB
Q5_K_L
Mixtral 8x7B Instruct v0.1
not listed
GPT OSS 20B
15.5 GB11.9 GB file
32K 16.0 GB · 128K 17.7 GB
BF16
Mixtral 8x7B Instruct v0.1
94.0 GB~93.4 GB file
32K 98.0 GB · 128K —
GPT OSS 20B
42.2 GB~41.8 GB file
32K 42.7 GB · 128K 44.4 GB
Mixtral 8x7B Instruct v0.1
Ollama · Q4_K_M
ollama run hf.co/TheBloke/Mixtral-8x7B-Instruct-v0.1-GGUF:Q4_K_MGPT OSS 20B
Ollama · Q4_K_M
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q4_K_M| Spec | Mixtral 8x7B Instruct v0.1 | GPT OSS 20B |
|---|---|---|
| Publisher | Mistral AI | OpenAI |
| Parameters | 46.7B | 20.9B |
| Active parameters | 13B per token | 4.2B per token |
| Architecture | MixtralForCausalLM | GptOssForCausalLM |
| Context | 32K tokens | 128K tokens |
| Modalities | Chat | Chat, Function-calling |
| License | Apache 2.0 | Apache 2.0 |
| Released | Dec 2023 | Aug 2025 |
Frequently Asked Questions
Which is better, Mixtral 8x7B Instruct v0.1 or GPT OSS 20B?
GPT OSS 20B leads llmrun Score 96.4 to 65.8; both fit 24 GB (Mixtral 8x7B Instruct v0.1 at IQ3_M, GPT OSS 20B at Q4_K_M); GPT OSS 20B is 2.1× faster on RTX 4090 (~190 vs ~92 tok/s). Choose GPT OSS 20B for higher llmrun Score and stronger at reasoning.
Which needs less VRAM, Mixtral 8x7B Instruct v0.1 or GPT OSS 20B?
At Q4_K_M, Mixtral 8x7B Instruct v0.1 requires 28.6 GB while GPT OSS 20B needs 12.9 GB, making GPT OSS 20B the easier model to run on your local setup.
Which is faster on RTX 4090?
Estimated decode speed on RTX 4090 is about 92 tok/s for Mixtral 8x7B Instruct v0.1 (IQ3_M) and 190 tok/s for GPT OSS 20B (Q4_K_M), so GPT OSS 20B is faster. Estimates, ±20%.
Can I run Mixtral 8x7B Instruct v0.1 or GPT OSS 20B on a 16 GB GPU?
Mixtral 8x7B Instruct v0.1 fits at IQ2_XXS; GPT OSS 20B fits at Q4_K_M.
Which has a longer context window, Mixtral 8x7B Instruct v0.1 or GPT OSS 20B?
Mixtral 8x7B Instruct v0.1 supports 32,768 tokens and GPT OSS 20B supports 131,072 tokens.