Is gpt-oss:20b good for coding?

In our benchmarks, gpt-oss:20b scores 63.3/100 for coding. It runs at about 77.7 tok/s, so if you want more speed, Nex-N2.5-mini-IQ3_XXS is faster (~91.4 tok/s) and still scores well for coding (81.6/100).

Is gpt-oss:20b good for agentic (tool-using) tasks?

In our benchmarks, gpt-oss:20b scores 77.7/100 for agentic workflows. It runs at about 77.7 tok/s, so if you want more speed, qwen3.8-flash-next-iq3_s is faster (~116.8 tok/s) and still scores well for agentic workflows (88.4/100).

How fast is gpt-oss:20b for local inference?

Across 35 community benchmark runs, gpt-oss:20b reaches up to 155.2 tok/s and averages 77.7 tok/s, with the fastest results on NVIDIA GeForce RTX 4070 Ti SUPER.

How much memory does gpt-oss:20b need?

The leanest observed configuration used about 382 MB of memory (quantizations tested: MXFP4, MXFP4_MOE).

Which tools have been used to run gpt-oss:20b?

Benchmarks were submitted using LM Studio, Ollama. Results are community-contributed and updated as new runs arrive.

Model benchmarks

gpt-oss:20b local LLM performance

As of October 2026, gpt-oss:20b runs at up to 155.2 tok/s for local inference (best of 35 community benchmark runs across 11 GPUs).

LM StudioOllamaMXFP4MXFP4_MOE

ShareReddit X

Model size

20.9B

Peak speed

155.2 tok/s

Average speed

77.7 tok/s

Avg PP

2083.9 tok/s

Min memory

382 MB

Max context

256,000 tokens

Avg output / run

9,771 tokens

Avg runtime / run

6m 17s

Avg quality

70.3

Benchmark runs

GPUs tested

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 35 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

Task	P5 (low)	Avg	P95 (high)
Overall	58.8	70.3	78.1
Agent Workflow	51.2	77.7	90.0
Code Generation	40.8	63.3	77.3
Role Play & Narrative	54.2	67.2	76.6
Research & Analysis	59.0	73.2	82.3

Performance by hardware and tool

Every hardware/tool/quantization combination gpt-oss:20b has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.

Hardware	Tool	Quant	Best tok/s	Avg tok/s	Memory	Context	Quality	Runs
NVIDIA GeForce RTX 4070 Ti SUPER	LM Studio	—	155.2 tok/s	151.7 tok/s	n/a	16,384 tokens	70.8	2
AMD Radeon RX 9070/9070 XT/9070 GRE	Ollama	MXFP4_MOE	132.4 tok/s	132.4 tok/s	10.9 GB	98,304 tokens	78.8	1
NVIDIA L40S	Ollama	MXFP4	120.7 tok/s	120.7 tok/s	12.0 GB	8,192 tokens	76.5	1
AMD Radeon RX 7900 XTX	LM Studio	—	111.1 tok/s	109.1 tok/s	n/a	256,000 tokens	67.0	9
AMD Radeon RX 7800 XT	Ollama	MXFP4	93.5 tok/s	89.7 tok/s	11.9 GB	65,536 tokens	73.5	3
AMD Radeon RX 7900 XTX	Ollama	MXFP4	70.6 tok/s	62.0 tok/s	13.1 GB	65,536 tokens	73.5	11
NVIDIA GeForce RTX 5080	Ollama	MXFP4	64.0 tok/s	35.0 tok/s	13.3 GB	65,536 tokens	62.4	2
Apple M5	LM Studio	—	45.3 tok/s	45.3 tok/s	n/a	8,192 tokens	74.0	1
NVIDIA GeForce RTX 5070	LM Studio	—	34.8 tok/s	34.8 tok/s	n/a	65,536 tokens	72.8	1
Apple M1 Pro	Ollama	MXFP4	31.1 tok/s	31.1 tok/s	11.9 GB	8,192 tokens	40.7	1
Apple M4	Ollama	MXFP4	25.3 tok/s	25.3 tok/s	382 MB	8,192 tokens	78.7	1
NVIDIA GeForce RTX 5070	Ollama	MXFP4	15.4 tok/s	15.4 tok/s	14.6 GB	65,536 tokens	68.2	1
Apple M3	Ollama	MXFP4	8.4 tok/s	8.4 tok/s	13.0 GB	8,192 tokens	74.3	1

Benchmark runs

All 35 gpt-oss:20b runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Open run page →

Frequently asked questions

Is gpt-oss:20b good for coding?: In our benchmarks, gpt-oss:20b scores 63.3/100 for coding. It runs at about 77.7 tok/s, so if you want more speed, Nex-N2.5-mini-IQ3_XXS is faster (~91.4 tok/s) and still scores well for coding (81.6/100).
Is gpt-oss:20b good for agentic (tool-using) tasks?: In our benchmarks, gpt-oss:20b scores 77.7/100 for agentic workflows. It runs at about 77.7 tok/s, so if you want more speed, qwen3.8-flash-next-iq3_s is faster (~116.8 tok/s) and still scores well for agentic workflows (88.4/100).
How fast is gpt-oss:20b for local inference?: Across 35 community benchmark runs, gpt-oss:20b reaches up to 155.2 tok/s and averages 77.7 tok/s, with the fastest results on NVIDIA GeForce RTX 4070 Ti SUPER.
How much memory does gpt-oss:20b need?: The leanest observed configuration used about 382 MB of memory (quantizations tested: MXFP4, MXFP4_MOE).
Which tools have been used to run gpt-oss:20b?: Benchmarks were submitted using LM Studio, Ollama. Results are community-contributed and updated as new runs arrive.