OpenAI

GPT-5 Mini

Type
image generation
Context
400K tokens
Max output
128K
Released
7 August 2025
Knowledge cutoff
May 2024
API string
openai/gpt-5-mini

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
qihang-ai $0.04 $0.29 31 Jul 2026
Poe $0.22 $1.8 31 Jul 2026
Jiekou.AI $0.225 $1.8 31 Jul 2026
Databricks databricks/databricks-gpt-5-mini $0.25 $2 31 Jul 2026
vivgrid $0.25 $2 31 Jul 2026
neon $0.25 $2 31 Jul 2026
perplexity-agent $0.25 $2 31 Jul 2026
302.AI $0.25 $2 31 Jul 2026
Abacus.AI $0.25 $2 31 Jul 2026
Azure Cognitive Services $0.25 $2 31 Jul 2026
Databricks databricks-gpt-5-mini $0.25 $2 31 Jul 2026
DigitalOcean Gradient AI $0.25 $2 1 Aug 2026
FastRouter $0.25 $2 31 Jul 2026
GitHub Copilot $0.25 $2 31 Jul 2026
Helicone $0.25 $2 31 Jul 2026
Kilo Code $0.25 $2 31 Jul 2026
LLM Gateway $0.25 $2 31 Jul 2026
Merge Gateway $0.25 $2 31 Jul 2026
Microsoft Foundry $0.25 $2 31 Jul 2026
NEAR AI $0.25 $2 31 Jul 2026
NanoGPT $0.25 $2 31 Jul 2026
OCI Generative AI $0.25 $2 31 Jul 2026
OpenAI $0.25 $2 31 Jul 2026
OpenRouter $0.25 $2 31 Jul 2026
OrcaRouter $0.25 $2 31 Jul 2026
Pioneer $0.25 $2 31 Jul 2026
Replicate $0.25 $2 31 Jul 2026
Requesty $0.25 $2 31 Jul 2026
SAP AI Core $0.25 $2 31 Jul 2026
Vercel $0.25 $2 31 Jul 2026
Snowflake $0.3 $1.2 1 Aug 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

A single provider appears several times if it sells this model under different identifiers and at different prices — for example, at different weight precision. The identifier is shown next to the name. Identical offers that arrived from two sources under different spellings are merged into one row.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score, programming 1,406.96 1,399–1,415 effort: high
Arena Score, expert questions 1,381.26 1,363–1,399 effort: high
Arena Score in French 1,380.89 1,349–1,413 effort: high
Arena Score, hard prompts 1,380.45 1,374–1,386 effort: high
Arena Score in English 1,379.92 1,374–1,386 effort: high
Arena Score, mathematics 1,376.7 1,361–1,392 effort: high
Arena Score, overall 1,373.84 1,369–1,378 effort: high
Arena Score, multi-turn dialogue 1,364.98 1,356–1,374 effort: high
Arena Score in Spanish 1,360.44 1,338–1,383 effort: high
Arena Score, instruction following 1,358.82 1,351–1,366 effort: high
Arena Score in Russian 1,358.03 1,343–1,373 effort: high
Arena Score, long queries 1,355.81 1,348–1,364 effort: high
Arena Score, creative writing 1,326.91 1,317–1,337 effort: high
Arena Score, understanding diagrams 1,203.89 1,189–1,218 effort: high
Arena Score, working with images 1,202.66 1,195–1,211 effort: high
Arena Score, text recognition in images 1,199.8 1,190–1,210 effort: high