Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (GPT-5 mini, 7 Aug 2025, shipped with nano; this rubric is for mini, nano is $0.05/$0.40 and the same 400k window): $0.25 in / $2.00 out per 1M = about $0.69 blended (developers.openai.com model page and The Decoder's price table); 400k context; text and image input. Reasoning, coding and speed null: no launch-era Artificial Analysis index or Terminal-Bench figure was opened (AA's later v4.0 re-run gives mini high 41, not launch). Only 2 usable reviews found: no tech-press hands-on of mini exists; the Respan comparison post-dates launch (Feb 2026).[2]
The three new GPT-5 models are clearly intended as a replacement for most of the rest of the OpenAI line-up.
The All-Rounder (GPT-5 mini): Best for general-purpose chat and high-volume reasoning.
Once users reach usage limits on GPT-5, the system automatically shifts to GPT-5 mini — a more lightweight but still highly capable fallback.
| GPT-5: Key characteristics, pricing and model card · Aug 7, 2025 | 0.40 |
|---|---|
| OpenAI launches GPT-5, nano, mini and Pro — not AGI, but capable of generating 'software-on-demand' | VentureBeat · Aug 7, 2025 | 0.75 |
| GPT-5 mini vs Gemini 3 Flash Preview vs Claude 4.5 Haiku · Feb 17, 2026 | 0.40 |
Photos from readers, kept apart from the pin's sourced media. Rate what you ate or saw, and on a restaurant pin tag the dishes right on your photo.
The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.
OpenAI's[2][3][4][5] developer post is dated August 7, 2025 and says "Today, we're releasing GPT-5 in our API platform" in three sizes, gpt-5, gpt-5-mini and gpt-5-nano; the API changelog lists "August 7, 2025: Released GPT-5 family of models in the API" and both model pages list a gpt-5-mini-2025-08-07 and gpt-5-nano-2025-08-07 snapshot.
Model page: gpt-5-mini, 400,000-token context with 272,000 max input and 128,000 max output, a 31 May 2024 knowledge cutoff, text and image input, $0.25 input, $0.025 cached and $2 output per 1 million tokens, snapshot gpt-5-mini-2025-08-07; the docs now point new cost-sensitive projects to GPT-5.6 Terra.[1]
Model page: gpt-5-nano, the fastest and lowest-cost GPT-5 variant for summarization and classification, 400,000-token context, a 31 May 2024 knowledge cutoff, $0.05 input, $0.005 cached and $0.40 output per 1 million tokens, snapshot gpt-5-nano-2025-08-07; the docs now recommend GPT-5.6 Luna for most new work.[1]
OpenAI's[1][2][3][5] API changelog: "August 7, 2025: Released GPT-5 family of models in the API", the entry that introduced gpt-5, gpt-5-mini and gpt-5-nano.
OpenAI's[1][2][3][4] main launch post: GPT-5 as a unified system with a router, and once usage limits are reached "a mini version of each model handles remaining queries"; free users who hit their GPT-5 limit move to GPT-5 mini, "a smaller, faster, and highly capable model".
Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.