GPT-4.1 fine-tuned
GPT-4.1, OpenAI's smartest non-reasoning model for instruction following and tool calling, listed as supporting the v1/fine-tuning endpoint; snapshot gpt-4.1-2025-04-14 is a supervised fine-tuning target.
text + image → text · made by OpenAI
Sold by 6 ways
| Seller | Lane | Rate |
|---|---|---|
| Microsoft Azure AI cloud | batch | $1.1per Mtok in$4.4per Mtok outraw.githubusercontent.com · read 2026-10-31 |
| OpenAI api | batch | $1.5per Mtok in$6per Mtok outdevelopers.openai.com · read 2026-08-24 |
| Microsoft Azure AI cloud | standard | $2.2per Mtok in$8.8per Mtok out$0.55per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| OpenAI api | fine-tuned | $3per Mtok in$12per Mtok outdevelopers.openai.com · read 2026-08-24 |
| Microsoft Azure AI cloud | priority | $3.85per Mtok in$15.4per Mtok out$0.96per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| ElectronHub aggregator | standard | $2per Mtok in$8per Mtok outapi.electronhub.ai · read 2026-10-31 |
Measured 29 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 8thof 15 | Cad eval — Epoch AI | Overall pass (%) | 0.42 |
| 18thof 42 | Fictionlivebench — Epoch AI | 120k token score | 0.625 |
| 22ndof 111 | Vectara Hallucination Leaderboard | Hallucination Rate | 5.6 |
| 34thof 35 | Swe bench verified — Epoch AI | mean_score | 48.542 |
| 35thof 108 | Math level 5 — Epoch AI | mean_score | 83.006 |
| 36thof 46 | Enigma eval — Epoch AI | Accuracy | 2.17 |
| 38thof 72 | Aider polyglot — Epoch AI | Percent correct | 52.4 |
| 50thof 56 | Humanity's Last Exam (Epoch AI replication) | Accuracy | 5.4 |
| 61stof 92 | Forecastbench — Epoch AI | Overall score | 62.3 |
| 67thof 72 | Frontiermath tier 4 — Epoch AI | mean_score | 0 |
| 68thof 101 | FrontierMath (Tiers 1-3) | mean_score | 5.517 |
| 69thof 93 | Simpleqa verified — Epoch AI | mean_score | 31.1 |
| 84thof 165 | LMArena · Vision | Score (Elo) | 1214 |
| 87thof 102 | Simplebench — Epoch AI | Score (AVG@5) | 0.27 |
| 107thof 132 | Ale bench — Epoch AI | Performance | 558.1 |
| 113thof 122 | Frontiermath tiers 1 3 v2 — Epoch AI | mean_score | 5.965 |
| 120thof 173 | Weirdml — Epoch AI | Accuracy | 39.04 |
| 140thof 414 | LMArena · Text | Score (Elo) | 1415 |
| 144thof 154 | Chess puzzles — Epoch AI | mean_score | 3 |
| 151stof 197 | Lmca — Epoch AI | Score | 25.61 |
| 164thof 235 | Dtbench — Epoch AI | Accuracy | 68.27 |
| 198thof 326 | GPQA Diamond | mean_score | 66.919 |
| 218thof 306 | OTIS Mock AIME 2024-2025 | mean_score | 38.333 |
| 252ndof 282 | ARC-AGI-2 (Public Eval) | Score (%) | 0 |
| 260thof 273 | ARC-AGI-1 (Public Eval) | Score (%) | 11.75 |
| 268thof 290 | Arc agi 2 — Epoch AI | Score | 0.42 |
| 276thof 298 | ARC-AGI-2 (Semi-Private) | Score (%) | 0.42 |
| 285thof 298 | ARC-AGI-1 (Semi-Private) | Score (%) | 5.5 |
| 298thof 309 | Arc agi — Epoch AI | Score | 5.5 |
About
GPT-4.1 fine-tuned — a text model from OpenAI, sold by 3 companies from $2 in and $8 out per million tokens, placed 22nd of 111 on Vectara Hallucination Leaderboard.
It takes text and images and returns text, with a context window of 1,047,576 tokens. The catalogue files it under reasoning and agents. Three companies sell it. The cheapest is $2 in and $8 out per million tokens at ElectronHub. Beside the standard rate there are batch, faster and separately routed lanes. It has been measured on 29 boards, and stands best at 22nd of 111 on Vectara Hallucination Leaderboard.
Every current figure
- Maker
- OpenAI
- Register
- model
- Takes
- text + image
- Returns
- text
- Context
- 1,047,576 tokens
- Licence
- not read
- Sellers
- 3
- Price
- $2 in and $8 out per million tokens — ElectronHub
- Boards
- 29
- Best place
- 22nd of 111 — Vectara Hallucination Leaderboard
Known as 2 names
gpt-4.1-ftgpt-4.1-2025-04-14