Skip to content

gpt-oss-120b (batch)

text → textHugging Face OpenRouter Cost calculator →

Best price · blended / 1M

$0.06

at Near AI

Input / output
$0.03 / $0.15
Providers
21
Spread
6.9×
7 days
+16.5%
30 days
0%
Context
128K
Prices per 1M tokens, blended: Near AI $0.06, Venice $0.06, CoreWeave $0.065, DekaLLM $0.0675, AkashML $0.0693, DeepInfra $0.0703, Crusoe $0.10, Sail Research $0.145, DigitalOcean $0.15, Google (global) $0.158, Mancer 2 $0.158, BaseTen $0.20, BaseTen (fp4/us) $0.20, Amazon Bedrock $0.263, Amazon Bedrock (eu-west-1) $0.263, Groq $0.263, Nebius $0.263, Parasail $0.263, SiliconFlow $0.263, Together $0.263, Parasail (fast) $0.30, SambaNova $0.343, PrimeIntellect $0.396, Cerebras $0.45.

Providers

  • Near AICheapest
    $0.06
    $0.03 in · $0.15 out
    −7.7% vs ref
    128K ctx99.9% uptimefp4
  • $0.06
    $0.03 in · $0.15 out
    −7.7% vs ref
    125K ctx100% uptimeint4
  • $0.065
    $0.03 in · $0.17 out
    ref vs ref
    128K ctx98.9% uptimefp4
  • $0.0675
    $0.03 in · $0.18 out
    +3.8% vs ref
    128K ctx96.5% uptimebf16
  • $0.0693
    $0.03 in · $0.187 out
    +6.5% vs ref
    128K ctx100% uptimefp4
  • $0.0703
    $0.037 in · $0.17 out
    +8.1% vs ref
    128K ctx43.5% uptimebf16
  • $0.10
    $0.05 in · $0.25 out
    +54% vs ref
    128K ctx100% uptimebf16
  • $0.145
    $0.06 in · $0.40 out
    +123% vs ref
    128K ctx100% uptimefp4
  • $0.15
    $0.06 in · $0.42 out
    +131% vs ref
    125K ctx100% uptime
  • global
    $0.158
    $0.09 in · $0.36 out
    +142% vs ref
    128K ctx99.7% uptime
  • $0.158
    $0.085 in · $0.375 out
    +142% vs ref
    128K ctx97.5% uptimefp8
  • $0.20
    $0.10 in · $0.50 out
    +208% vs ref
    128K ctx100% uptimefp4
  • fp4/us
    $0.20
    $0.10 in · $0.50 out
    +208% vs ref
    128K ctx99.9% uptimefp4
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx100% uptime
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx99.9% uptime
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx99.6% uptime
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx100% uptimefp4
  • $0.263
    $0.10 in · $0.75 out
    +304% vs ref
    128K ctx100% uptimefp4
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx88.2% uptimefp8
  • $0.263
    $0.15 in · $0.60 out
    +304% vs ref
    128K ctx99.9% uptime
  • $0.30
    $0.15 in · $0.75 out
    +362% vs ref
    128K ctx100% uptimefp4
  • $0.343
    $0.14 in · $0.95 out
    +427% vs ref
    128K ctx98.6% uptime
  • $0.396
    $0.308 in · $0.661 out
    +510% vs ref
    128K ctx98.4% uptimefp4
  • $0.45
    $0.35 in · $0.75 out
    +592% vs ref
    128K ctx100% uptimefp16
Blended = (3 × input + output) / 4. Ref $0.065 = median of cheapest 20%.

Price history

USD per 1M tokens · each step is a price change

Prices in USD / 1M tokens.

Other pricing

Offered byCache read
Near AI$0.015
Venice, CoreWeave +3$0.03
DeepInfra$0.0185
Crusoe$0.05
DigitalOcean$0.012
Mancer 2$0.085
BaseTen, BaseTen (fp4/us) +1$0.10
Groq, SiliconFlow$0.075
Parasail$0.055
PrimeIntellect$0.278
Cerebras$0.35

Changes

  • Cache read price,
    $0.315→$0.278−11.7%
  • Output price,
    $0.75→$0.661−11.9%
  • Input price,
    $0.35→$0.308−12.0%
  • Cache read price,
    $0.095→$0.085−10.5%
  • Input price,
    $0.095→$0.085−10.5%
  • Cache read price,
    $0.266→$0.315+18.4%
  • Output price,
    $0.634→$0.75+18.3%
  • Input price,
    $0.296→$0.35+18.2%
  • Cache read price,
    $0.315→$0.266−15.6%
  • Output price,
    $0.75→$0.634−15.5%
  • Input price,
    $0.35→$0.296−15.4%
  • Cache read price,
    $0.123→$0.315+156%
  • Output price,
    $0.293→$0.75+156%
  • Input price,
    $0.137→$0.35+155%
  • Cache read price,
    $0.069→$0.123+78.3%

About

The cheapest provider for gpt-oss-120b (batch) is Near AI at $0.06 per 1M tokens (blended), charging $0.03 per 1M input and $0.15 per 1M output tokens. 21 providers serve it; the most expensive charges 6.9× the typical price. Context window: 128K tokens. Prices tracked since Oct 3, 2026.

gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...
Released
Aug 5, 2025
First tracked
Oct 3, 2026
Canonical slug
openai/gpt-oss-120b
Weights
openai/gpt-oss-120b