GPT 6.1 Sol Review: 1M Context, Five Effort Levels, Tool Calls
Pick GPT 6.1 Sol for hard coding and analysis that needs a million-token window and a real reasoning dial; so far we have run only launch checks.
Freedom Score
Not yet tested
- Price
- $4.00 per 1M mixed tokens (3 input : 1 output)
- Output speed
- 26 tok/s
- Success rate
- Collecting data
GPT 6.1 Sol is the larger of OpenAI's two GPT 6.1 models, released on 29 September 2026. OpenAI positions it for complex coding, computer use and professional work, and describes it as near-Astra performance at a lower cost. On our API it is one OpenAI-compatible chat endpoint that takes text and images and returns text, with a 1,050,000-token window and replies up to 128,000 tokens.
- SpicyAPI Engineering
- Tested Oct 28, 2026
- Methodology v3.3 →
- We sell these models; scores ignore price →
The verdict
This is a launch-day review. On 28 October 2026 our platform ran a real request through the endpoint as part of launch: one neutral message, answered without streaming at low reasoning effort. It returned a complete answer. We have not run a quality blind test, a long-document test or a content-level test, and this page does not score the model; where a result is missing we say so below.
Choose Sol when a task is hard enough that a missed detail is expensive: a bug that crosses several files, an analysis with many moving parts, an agent that is allowed to change things. For routine traffic, the cheaper GPT 6.1 Terra shares the same window and reply ceiling.
Key facts
- Tested: one real launch request on 28 Oct 2026 (neutral prompt, low reasoning effort, no streaming). No scored runs yet.
- Price: $2 per million input tokens, $0.10 cached and $10 output; prompts over 272,000 tokens are billed on a higher tier for the whole request. The live price is on the model page.
- Limits: 1,050,000-token context, replies up to 128,000 tokens (default 4,096, reasoning included), text and image input, 256 messages and 128 tool definitions per request.
- Reasoning: reasoning_effort takes low, medium, high, xhigh or max (default medium); there is no none.
- Sources: scoring rules are on our review methodology page, full parameters on the GPT 6.1 Sol model page.
What we liked and what held it back
What we liked
- A million-token window with a 128,000-token reply ceiling. A repository, a specification and the question fit in one request, and a long answer comes back in one piece.
- Five effort levels on one field. Low for quick passes, max for cases that justify the tokens, and the switch is one string on the same request body.
- Tool calling is on the endpoint. Up to 128 function definitions per request, with toolchoice and paralleltool_calls accepted; we have not graded tool-use quality yet.
- Sampling stays adjustable. temperature (0 to 2), top_p and up to four stop sequences are available.
- Same request shape as Terra. Moving traffic between the two models is a model-name change, not a rewrite.
What held it back
- No none setting. Even the lightest request gets a short reasoning pass; the cheaper Terra is the GPT 6.1 model that accepts none.
- A price cliff at 272,000 tokens. A prompt over that size is billed on a higher tier for the whole request, not just the excess, so send a full codebase only when the task needs it.
- Reasoning is paid for out of max_tokens. The default cap is 4,096; a small cap at a high effort can be spent mostly on thinking before the visible answer starts.
- Images are capped at 16,384 input tokens each. Enough for a screenshot or diagram; a dense scanned page may need a smaller crop, and Terra allows 37,000 per image.
- No audio or video input, and no quality numbers of our own yet.
What we tested
On 28 October 2026 our platform sent GPT 6.1 Sol one real request as part of launch: a single neutral user message, answered without streaming at low reasoning effort with a 1,024-token cap. It returned a complete answer, and the request was priced at the published rate. The endpoint also passed our pre-launch checks of its price table and request limits.
That is a working-endpoint check, not a benchmark. We have not compared Sol's answers with another model, tested streaming or tool calls beyond what the endpoint documents, measured speed under load, or tried the 272,000-token tier. We plan to add scored runs; until then, treat the capability claims on this page as OpenAI's own and the limits as what our endpoint accepts.
Prompts to start from
These starting prompts come from the model page's published examples. We have not graded the outputs, so they are a place to begin rather than prompts proven on this model.
1. Review a change with a defined target. Name the failure you are hunting and ask for evidence.
Review a retry-loop change and identify where duplicate billing could occur. Work from the material I provide, cite the relevant lines, distinguish facts from assumptions, and say what information is missing before proposing changes.Run this prompt on GPT 6.1 Sol →
2. A tool step. Confirms your function-calling loop end to end with a trivial function.
Call report_status with status ready.Settings that matter: start at medium effort and step up only where a wrong answer is costly; leave several thousand tokens of max_tokens headroom above the visible answer; put reference material first and the question last.
How far it writes
We have not run our content-level test on GPT 6.1 Sol, so we have no result to report and the endpoint carries no Unrestricted label. The model page states the position plainly: content decisions come from the model, its behaviour can vary between versions, and no guarantee of unrestricted output is made.
Not tested yet: the full content-level ladder, multi-turn behaviour over a long session, languages other than English, and how the 272,000-token tier behaves in practice.
Every character in a request should be an adult, keep real, identifiable people out of it, and you are responsible for the law where you publish.
What the community says
The launch thread drew a split reaction. One reader summed up the pricing news as "Shots fired, half the price of Opus 5.5." Another welcomed it but wondered about the motive: "This is great. But maybe part of the motivation is that 6-Sol wasn't as good as initially advertised so they needed to tweak it."
Alternatives
- The same family at half the price? GPT 6.1 Terra charges $1 per million input tokens and $5 output, accepts reasoning_effort none and allows larger images; Sol is the pick for the hardest tasks.
- A different 1M-class model with audio input? Gemini 4 Argon reads text, images and audio, with replies capped at 65,536 tokens.
- A different publisher's flagship? Claude Opus 5.5 is the Anthropic option on the same API; compare it on your own prompts.
How GPT 6.1 Sol earns its Spicy Index
Each dimension below links to its full workings: the table row every point came from and why points were lost.
- Public leaderboards 40% of quality
- Not yet tested
- Role-play 15% of quality
- Not yet tested
- Creative writing 15% of quality
- 72.7
- Instruction following 15% of quality
- 100.0
- Long-context memory 15% of quality
- Not yet tested
- Capability 10.5% weight
- 52.0
- Freedom Score 30% weight
- Not yet tested
No exact match on a tracked public leaderboard yet. How it works
Our role-play suite hasn’t run on this model yet. How it works
Our long-context suite hasn’t run on this model yet. How it works
Freedom evidence is untested or insufficient. The Index retains the original quality and capability score, is marked preliminary and still receives a rank. Missing evidence is not zero. How it works
Freedom evidence is untested or insufficient. The Index retains the original quality and capability score, is marked preliminary and still receives a rank. Missing evidence is not zero.
Open the full test report →Capability, item by item
Points from the public scoring table. Upscaled tiers and limits we set ourselves don’t count.
- Context window (creator-published)Not declared
- 0.0 / 30
- Input modalities on SpicyAPIText, plus images
- 12.0 / 25
- Tool useTools and tool_choice
- 15.0 / 15
- Structured outputAccepted, schema adherence verified
- 15.0 / 15
- Reasoning control5 levels: low · medium · high · xhigh · max
- 10.0 / 15
What GPT 6.1 Sol can do on SpicyAPI
Straight from the live model specification on SpicyAPI. Some creators offer more in their own apps; this is what you can call through our API today.
| Feature | Chat |
|---|---|
| Context window | 1,050,000 tokens |
| Max output per request | 128,000 tokens |
| Reasoning levels | low · medium · high · xhigh · max |
| Tool calling | Yes |
| Structured output | Yes |
How long it really takes
Measured on real jobs over the last 30 days, from request to finished result. Times include our own test runs.
- Output speed
- 26 tok/sIncludes time to first token
- Success rate
- Collecting data
A time appears once at least 10 completed jobs are in. Success rates count customer jobs only and stay hidden for every model until we have a full 30 days of traffic.
What people say about GPT 6.1 Sol
1 positive · 1 mixed · 0 negative
Shots fired, half the price of Opus 5.5.
hacker-news· Sep 29, 2026· PositiveSource
This is great. But maybe part of the motivation is that 6-Sol wasn't as good as initially advertised so they needed to tweak it.
hacker-news· Sep 29, 2026· MixedSource
Quotes are verbatim from public posts, linked to the original. They are not ratings and don’t feed any score.
Models to consider instead of GPT 6.1 Sol
OpenAI
GPT 6.1 Terra
The same family at half the price?
- Spicy Index
- 79.9−1.3
- Freedom Score
- —
- Price
- $2.00−50%/ 1M tokens · 3:1 input/output
Anthropic
Claude Opus 5.5
A different publisher's flagship?
- Spicy Index
- 82.5+1.3
- Freedom Score
- 88.9
- Price
- $7.20+80%/ 1M tokens · 3:1 input/output
Google
Gemini 4 Argon
A different 1M-class model with audio input?
- Spicy Index
- 76.7−4.5
- Freedom Score
- —
- Price
- $8.00+100%/ 1M tokens · 3:1 input/output
What do you think?
Was this review useful?
Log in to vote.
Comments
Comments are checked automatically and by our team. We remove spam, offensive content, personal attacks and anything that names a real person without consent. Report a comment and we review it within one business day.
Tried this model? Share what worked. You need a free account to comment, so we can keep spam out.
GPT 6.1 Sol FAQ
Try GPT 6.1 Sol yourself
Set every parameter in the Playground and see the exact price before you run it.
$4.00 per 1M mixed tokens (3 input : 1 output)
More reviews
See the leaderboards →GPT 6.1 Terra
Pick GPT 6.1 Terra for high-volume screens, documents and writing where reasoning_effort none helps; so far we have run only launch checks, no scored tests.
- Spicy Index
- 79.9
- Freedom Score
- —
Tested Oct 28, 2026
Read the reviewKimi K3.1 Pro
Pick Kimi K3.1 Pro for long-context work with screenshots and clips when you accept always-on thinking; so far we have run only launch checks, no scored tests.
- Spicy Index
- 80.4
- Freedom Score
- —
Tested Oct 28, 2026
Read the review
Grok 4.3
Pick Grok 4.3 for a huge context window, a genuine reasoning off switch and uncensored writing at a low price; skip it if you need file uploads.
Tested Sep 24, 2026
Read the review