Compare / Llama 4 Maverick vs gpt-oss-20b

Llama 4 Maverick vs gpt-oss-20b

Which is cheaper, which is more capable, and which does better at each kind of work, from 23 results on 3 sources that measured both models.

MetaLlama 4 Maverick Released 5 Apr 2025 · $0.35 in · $1.00 out
OpenAIgpt-oss-20b Released 5 Aug 2025 · $0.030 in · $0.14 out
Llama 4 Maverick wins 74 too close to callgpt-oss-20b wins 12

Of the 23 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 19 November 2026 (catalogue release 2026-11-19-3608133e8331). See also Meta vs OpenAI, every model from both labs.

Llama 4 Maverick or gpt-oss-20b? The short answer

They score about the same, so pick gpt-oss-20b: it costs 8.9 times less per token.

  • gpt-oss-20b is cheaper: $0.058 against $0.51 per million tokens, 8.9 times less.
  • They score about the same on the Artificial Analysis Intelligence Index (9.3 and 10.0).
  • gpt-oss-20b writes faster: 277 tokens a second against 46.
  • gpt-oss-20b is newer: released 5 Aug 2025, against 5 Apr 2025 for Llama 4 Maverick.
  • On third-party benchmarks, on the results both models have: gpt-oss-20b does better for agents and tool use, reasoning and knowledge and writing.

Head to head

9.3
Intelligence Index
10.0
$0.51
Price per million tokens, blendedlower is better
$0.058
46
Output speed, tokens a second
277
524,288
Context window, tokens
131,072

Every fact side by side

Llama 4 Maverickgpt-oss-20b
Price per million tokens$0.35 in · $1.00 out$0.030 in · $0.14 out
Blended (3 in : 1 out)$0.51$0.058
Cheapest host, blended$0.51 (Parasail, FP8)$0.022 (PrimeIntellect, FP4)
Intelligence Index9.310.0
Spring Prompt overall––
Output speed, tokens a second46277
Context window524,288 tokens131,072 tokens
Released5 Apr 20255 Aug 2025 (newer)
WeightsOpenOpen
Developer based inthe USthe US

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 143 of 159 130 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 142 of 187 132 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 152 of 169 155 of 169

From our research

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureLlama 4 Maverickgpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 220 · 237 of 387 measured 1,327 1,318
Business, management and finance · TextArena rating, higher is better 222 · 213 of 380 measured 1,317 1,323
Creative writing · TextArena rating, higher is better 199 · 275 of 385 measured 1,307 1,240
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 185 · 176 of 249 measured 9.3 10.0low reasoning
GPQA Diamond% of questions, higher is better 165 · 159 of 232 measured 67.1% 68.8%high reasoning
Humanity's Last Exam% of questions, higher is better 201 · 153 of 250 measured 4.9% 11.0%high reasoning
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 186 · 182 of 230 measured 24.0 24.6low reasoning
Requested-length error · UGI% off the requested word count, lower is better 57 · 221 of 230 measured 13.0% 98.0%low reasoning
Style adherence · UGIscore from 0 to 1, higher is better 45 · 187 of 230 measured 0.37 0.31medium reasoning
Show all 23 results (14 more)
MeasureLlama 4 Maverickgpt-oss-20b
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 211 · 220 of 338 measured 1,326 1,314
Instruction following · TextArena rating, higher is better 222 · 257 of 387 measured 1,314 1,281
Writing, literature and language · TextArena rating, higher is better 210 · 261 of 386 measured 1,312 1,275
Artificial Analysis reported by Artificial Analysis · 11 results
Terminal-Bench 2.1% of tasks, higher is better 107 · 97 of 131 measured 7.9% 13.9%high reasoning
Terminal-Bench 4.0% of tasks, higher is better 72 · 72 of 120 measured 0.0% 0.0%high reasoning
τ-bench banking% of tasks, higher is better 114 · 90 of 121 measured 3.7% 7.0%high reasoning
Artificial Analysis Coding Indexindex score, higher is better 106 · 100 of 130 measured 16.3 20.7high reasoning
IFBench% of instructions, higher is better 122 · 69 of 195 measured 43.0% 65.1%high reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 162 · 184 of 241 measured 50.0% 34.7%high reasoning
Output speedtokens per second, higher is better 70 · 6 of 78 measured 46.4 277low reasoning
SciCode% of problems, higher is better 106 · 90 of 121 measured 31.7% 38.9%high reasoning
Terminal-Bench Hard% of tasks, higher is better 140 · 127 of 190 measured 6.8% 10.6%high reasoning
Time to first answer tokenseconds, lower is better 15 · 44 of 78 measured 0.65 7.62low reasoning
τ²-bench telecom% of tasks, higher is better 171 · 96 of 193 measured 17.8% 60.2%high reasoning

Quick answers

Is Llama 4 Maverick better than gpt-oss-20b?

They score about the same, so pick gpt-oss-20b: it costs 8.9 times less per token.

Which is cheaper, Llama 4 Maverick or gpt-oss-20b?

gpt-oss-20b is cheaper: $0.058 against $0.51 per million tokens, 8.9 times less.

Which is faster, Llama 4 Maverick or gpt-oss-20b?

gpt-oss-20b writes about 277 tokens a second on its usual API, against 46 for Llama 4 Maverick (Artificial Analysis).

Which is better for agents and tool use, Llama 4 Maverick or gpt-oss-20b?

gpt-oss-20b ranks 130 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 143 for Llama 4 Maverick.

Which is better for reasoning and knowledge, Llama 4 Maverick or gpt-oss-20b?

gpt-oss-20b ranks 132 of 187 models on sale for hard questions across science, maths and general knowledge, against 142 for Llama 4 Maverick.

Which is better for writing, Llama 4 Maverick or gpt-oss-20b?

Llama 4 Maverick ranks 152 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 155 for gpt-oss-20b.

Which has the bigger context window, Llama 4 Maverick or gpt-oss-20b?

Llama 4 Maverick: 524,288 tokens, against 131,072 for gpt-oss-20b.

Run Llama 4 Maverick and gpt-oss-20b on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.