GPT-6.1 Sol
OpenAI's DevDay 2026 model: near-Astra capability for agentic coding and professional work at one fifth of Astra's price.
What it is
GPT-6.1 Sol was released on September 29, 2026, exactly one week after GPT-6 Sol. It keeps GPT-6 Sol's $2 / $10 list price, halves the cached-input price to $0.10, and OpenAI says it delivers nearly the same level of intelligence as GPT-6 Astra for agentic coding, computer use and professional work. It is the model behind Codex for paid ChatGPT tiers.
Pricing
| Input | $2.00 |
| Output | $10.00 |
| Cached input (cache read) | $0.100 |
| Cache write | $2.50 |
| Batch input | $1.00 |
| Batch output | $5.00 |
| Fast / priority input | $4.00 |
| Fast / priority output | $20.00 |
- Long context (>272K input): 2x input and cache rates, 1.5x output ($4 / $15).
- Batch and Flex are 50% of Standard; Fast mode is 2x.
- Ultrafast tier announced for Astra first; no published Sol 6.1 Ultrafast price as of October 8, 2026.
- Regional processing adds a 10% premium where available.
Capability highlights
- OpenAI positions it as close to GPT-6 Astra on agentic coding, computer use and professional work at 1/5 the token price (vendor claim).
- Multi-agent support in beta from launch (OpenAI API changelog).
- Responses API tools: web search, file search, code interpreter, hosted shell, apply_patch, computer use, MCP, tool search.
- Chat Completions endpoint works but without tool calling; use the Responses API for agents.
Modalities: text, image → text. Reasoning model; effort low, medium (default), high, xhigh, max. none and minimal not supported. Prompts over 272K input tokens are billed at long-context rates for the whole request.
How to use it
API: OpenAI API model ID gpt-6.1-sol (Responses, Chat Completions, Batch). Also generally available on Amazon Bedrock since September 29, 2026.
Versus GPT-6 Sol
Same $2 / $10; cached input halved from $0.20 to $0.10; knowledge cutoff moved from April 20 to April 30, 2026.
| GPT-6 Sol | GPT-6.1 Sol | |
|---|---|---|
| Input / 1M | $2.00 | $2.00 |
| Output / 1M | $10.00 | $10.00 |
| Cached input / 1M | $0.200 | $0.100 |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 128K |
Published benchmark numbers
Vendor-reported, quoted from the linked announcement. Not independently verified; different vendors run benchmarks under different settings.
| Benchmark | GPT-6.1 Sol | Comparison in the same source | Source |
|---|---|---|---|
| Factual-error rate on hard prompts (low effort) OpenAI figure as reported by TechCrunch; OpenAI's launch page was not readable by our fetcher. | 7.7% | GPT-6 Sol: 11.4% | TechCrunch |
Comparisons
Sources
Every number above comes from one of these pages, read on the date shown. Searched for: 6.1 sol, gpt sol 6.1, gpt 6.1 sol, codex 6.1 sol, gpt 6.1 sol pricing.
- OpenAI model page: gpt-6.1-sol read 2026-10-08
context, max output, cutoff, pricing, features, endpoints - OpenAI API pricing read 2026-10-08
standard, batch, flex, fast, long-context rates - OpenAI API changelog (Sep 29, 2026 entry) read 2026-10-08
release date, multi-agent beta - AWS What's New: GPT-6.1 Sol on Amazon Bedrock read 2026-10-08
Bedrock availability - TechCrunch: OpenAI launches GPT-6.1 Sol (secondary; quotes OpenAI) read 2026-10-08
availability by tier, OpenAI's capability claims