GPT-6.1 Sol Is Live on Smart AIPI — Near-Astra Intelligence at a Fifth of the Price

OpenAI has released GPT-6.1 Sol, an upgrade to GPT-6 Sol that nearly matches GPT-6 Astra's intelligence on agentic coding, computer use, and professional work — at one-fifth of Astra's standard input and output token prices. It's available in the API as gpt-6.1-sol, and it's on Smart AIPI right now at a quarter of OpenAI's price.

Launch posts are marketing. This one isn't — we sell access to the model, so the useful thing we can do is tell you exactly where it wins, where Astra still wins, and what it will actually cost you. All numbers below come from OpenAI's published evaluations.

Where GPT-6.1 Sol lands in the lineup

On Smart AIPI you pay 25% of those numbers. The headline: GPT-6.1 Sol delivers substantial improvements over GPT-6 Sol across complex professional tasks — writing and debugging code, understanding documents, and executing multi-step business workflows — and on several evaluations it approaches Astra's performance at a fraction of the cost.

The results that matter for developers

Coding

On DeepSWE v1.1, which evaluates complex software-engineering tasks in real codebases, GPT-6.1 Sol matches GPT-6 Astra at roughly one-fifth of the cost, while eclipsing GPT-6 Sol's best score by 6.4 percentage points at a lower reasoning effort and cost. That's the rare upgrade that is both better and cheaper to run.

Professional work

On GDP.pdf — real prompts about complex PDFs (tables, charts, diagrams, fine print) from finance, healthcare, legal, and seven other domains — GPT-6.1 Sol scores higher than Opus 5.5 with fallbacks at less than half the cost per task, and approaches Astra's state-of-the-art at roughly one-fifth the cost. On AutomationBench (multi-step business workflows across 47 tools), it scores 2.2 points above Opus 5.5 at medium reasoning effort for about a third of the cost — up 4.8 points from GPT-6 Sol at the same setting.

Computer use

On OSWorld 2.0's offline set, GPT-6.1 Sol outperforms GPT-6 Sol by seven percentage points at maximum reasoning effort for less than half the cost, and comes within 2.1 points of Astra at roughly one-seventh the cost per task. Astra is still the strongest computer-use model; GPT-6.1 Sol gets most of the way there for a fraction of the spend.

Scientific research

On Terminal-Bench Science 0.1 (data analysis, simulation, theorem proving), GPT-6.1 Sol more than doubles GPT-6 Sol's score at maximum effort for under half the cost. At max effort it costs $5.47 per task on average, versus $23.21 for Opus 5.5 and $23.80 for Astra — over 75% cheaper than either. Astra still tops the benchmark at 68.1% and remains the pick for the hardest research tasks.

Factuality

On difficult, error-inducing prompts, GPT-6.1 Sol's biggest gain is at low reasoning effort, where it cuts the share of answers with a factual error from 11.4% to 7.7% — about a 32% reduction. Across the tested settings its error rate stays within 1.9 points of Astra at under a fifth of the cost. (These are deliberately hard prompts and don't reflect typical usage, where factual errors are rarer.)

Alignment: measurably closer to Astra

GPT-6.1 Sol shows substantial improvements over GPT-6 Sol in OpenAI's alignment evaluations — it's more transparent about its limitations and more reliable at respecting user intent and safety constraints. One concrete example, disclosing when a search tool is broken instead of guessing (lower is better):

It also showed no attempts to bypass an automated safety reviewer, matching both GPT-6 Astra and GPT-6 Sol. Full details are in OpenAI's GPT-6.1 Sol system card addendum.

The cached-input story

The number that quietly matters most for agents: cached input is $0.10 per 1M at OpenAI list — 95% cheaper than standard input, and half the cached rate of GPT-6 Sol. On Smart AIPI that lands at $0.025 per 1M cached input. If your agent reuses a large system prompt, tool schema, or document context across requests, that context is nearly free to re-send. Combined with the 75% Smart AIPI discount, GPT-6.1 Sol is one of the cheapest ways to run a capable, context-heavy agent.

This walkthrough breaks down why the halved cache-read price changes the economics of context-heavy agents — and makes the case for defaulting to GPT-6.1 Sol going forward:

The takeaway lines up with the numbers above: with cache reads at half the price and near-Astra quality, GPT-6.1 Sol is the sensible new default for most agent workloads — reach for Astra only on the hardest tasks.

Pricing: 75% off, same model, same API

Same weights, same OpenAI-compatible API shape, same tools. The only difference is the invoice.

How to use GPT-6.1 Sol

Chat Completions API

curl https://api.smartaipi.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6.1-sol",
    "messages": [{"role": "user", "content": "Refactor this module and add tests"}]
  }'

Responses API with high reasoning

curl https://api.smartaipi.com/v1/responses \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6.1-sol",
    "input": "Triage these open issues and open PRs to fix the top three",
    "reasoning": {"effort": "high", "summary": "auto"}
  }'

Codex CLI, Cursor, Cline, Roo Code, OpenCode, Aider

OPENAI_BASE_URL=https://api.smartaipi.com/v1
OPENAI_API_KEY=sk-your-smart-aipi-key
OPENAI_MODEL=gpt-6.1-sol

Reasoning effort supports low, medium, high, and xhigh. GPT-6.1 Sol scales well with effort — several of the results above are at maximum effort, where it still costs a fraction of Astra per task. A GPT-6.1 Sol Ultrafast variant with up to 8x faster token generation is coming.

Which model should you actually run?

• gpt-6-astra — the most intelligent model, for the best results and the hardest tasks: the toughest research, long-horizon computer use, anything where finishing unattended is the point.
• gpt-6.1-sol — near-Astra intelligence at a fifth of the price. The new default for serious agentic coding, document work, business automation, and science where you want most of Astra's quality without Astra's bill.
• gpt-6-sol — still a strong, cheaper-cached option; gpt-6.1-sol supersedes it on quality at the same input/output price.
• gpt-6-luna — fast, efficient everyday work at scale: bulk tasks, classification, summarization, high-throughput pipelines.

All four run on the same endpoint and the same key. Switching is a one-line change, so benchmark them on your workload rather than trusting anyone's chart — including this one.

Get started

1. Sign up at smartaipi.com/signup (free credits, no credit card)
2. Create an API key in the dashboard
3. Set your base URL to https://api.smartaipi.com/v1
4. Use model gpt-6.1-sol

Near-Astra intelligence, in production, at a fifth of the price — and a quarter of that again on Smart AIPI.