GPT-6.2 Sol API

GPT-6.2 Sol launched on 2026-11-17 alongside GPT-6.2 Astra. OpenAI calls it its go-to model for everyday work, from coding to powering agents at scale, and says many OpenAI employees use it as their default. It keeps GPT-6.1 Sol’s price of $2 input / $10 output per 1M tokens, with cached input at $0.10; on OpenAI’s FrontierCode 2.0 it lands within 0.5 points of Astra at about a fifth of the cost per task. This page covers the benchmarks and third-party evaluations, how to choose between it, GPT-6.1 Sol, Astra and Claude, Ultrafast and Codex limits, what to change when migrating, and Wokey pricing.

Price check: official, OpenRouter and Wokey

USD per million tokens. Official and OpenRouter rates were checked by hand on 2026-11-18, excluding tax; the Wokey rate is read from the live price list.

ChannelInputCache readOutput
Official API$2$0.1$10
OpenRouter$2$0.1$10
Wokey$0.18$0.009$0.9

· Official pricing page · OpenRouter endpoint list

GPT-6.2 Sol at a glance

Release
2026-11-17, together with GPT-6.2 Astra; available the same day in the API, ChatGPT Work and Codex, and listed on OpenRouter (OpenAI, Azure and Amazon Bedrock endpoints)
Positioning
An everyday model that balances intelligence and cost, built for multistep agentic workflows and clear, natural writing, and easy to collaborate with. OpenAI says it is smarter and more token-efficient than GPT-6.1 Sol and completes more tasks on the same paid plan; new ChatGPT Work chats default to Sol
Context and output
1,050,000-token context (922K max input) and 128K max output; text and image input, text output, no audio or video; knowledge cutoff 2026-06-01 (GPT-6.1 Sol: 2026-04-30)
Reasoning effort
The API rejects none. OpenAI suggests starting new integrations at medium and moving none or minimal to low; the launch charts and Codex offer low, medium, high, xhigh and max, and the docs do not yet state a default
Official pricing
$2 input / $10 output per 1M tokens, cached input $0.10, cache writes $2.50 (30-minute retention) or $5 (12 hours); requests over 272K input tokens are billed at 2× input and cache and 1.5× output for the whole request ($4 / $0.20 / $15); Batch 50% off, Fast 2×, Ultrafast 3× through 2027-02-28 and 6× after
Ultrafast
OpenAI says Ultrafast generates up to 4.5× faster than Standard. It is in the API from launch (US data residency only); in ChatGPT, Pro $200 gets Ultrafast for the first time on Sol, and Pro $500 gets it on both models
API and tools
Chat Completions, Responses and Batch; tool calling requires Responses. Built-in tools: web_search, file_search, image_generation, code_interpreter, hosted_shell, apply_patch, skills, computer_use, mcp and tool_search. No fine-tuning, Realtime or Assistants, and no dated snapshot
Model ID and limits
gpt-6.2-sol; top tier 15,000 RPM / 40M TPM; US and EU data residency, with a 10% uplift on regional endpoints

· Sources: OpenAI · OpenAI API · GPT-6.2 Sol · Latest model guide · Codex pricing · Codex models · Artificial Analysis · Vals AI · VentureBeat

Official benchmarks: GPT-6.2 Sol vs GPT-6.1 Sol vs GPT-6.2 Astra vs Claude Opus 5.5 vs Fable 5.2

From the nine comparison charts in OpenAI’s launch post. These are vendor-run results, each at the highest effort OpenAI plotted for the model.

BenchmarkGPT-6.2 SolGPT-6.1 SolGPT-6.2 AstraClaude Opus 5.5Claude Fable 5.2
Agents’ Last Exam v1.173.0%72.5%77.6%71.6%72.8%
Terminal-Bench-Science 0.168.0%57.0%77.1%63.3%73.3%
GDPval v2 (internal)46.9%41.5%50.4%43.5%44.8%
FrontierCode 2.0 (Cognition)43.8%39.0%44.2%45.5%45.7%
AutomationBench 1.0.641.1%36.1%52.4%42.5%51.4%
BenchCAD 2.1 Core73.1%52.6%82.5%65.3%83.5%
Harvey LAB-AA v1.111.1%6.9%15.0%4.2%7.2%
Vals Science Index v157.8%59.6%68.4%58.7%63.7%
LifeVal83.8%79.5%86.1%——

GPT-6.2 Sol beats GPT-6.1 Sol on eight of the nine; the exception is the Vals Science Index, where even its xhigh score of 58.5% stays below GPT-6.1 Sol’s 59.6%. Fable 5.2’s Terminal-Bench-Science score is Artificial Analysis’ mini-swe-agent run, and its Vals Science Index score is at high effort. On FrontierCode 2.0, GPT-6.2 Sol costs $1.29 per task against $5.97 for GPT-6.2 Astra; on Agents’ Last Exam, Sol costs $1.32 per task against $17.57 for Opus 5.5.

Which to use: GPT-6.2 Sol, GPT-6.1 Sol, GPT-6.2 Astra or Claude

Versus GPT-6.1 Sol

Same price, and the knowledge cutoff moves from 2026-04-30 to 2026-06-01. It leads on eight of OpenAI’s nine comparisons, gains 5.3 points on GDPval v2 and goes from 57.0% to 68.0% on Terminal-Bench-Science; the Artificial Analysis Intelligence Index rises from 55 to 57. AA’s Coding Agent Index (Codex harness, max effort), however, drops 3 points, mostly on SWE-Atlas-QnA. With no price change, use 6.2 Sol for new work, and compare first if your workflow is mostly code Q&A.

Versus GPT-6.2 Astra

Astra costs four times as much ($8 / $40). On OpenAI’s FrontierCode 2.0 at max effort, Sol scores 43.8% against Astra’s 44.2% at about a fifth of the cost per task, but Astra is clearly stronger on Agents’ Last Exam (77.6% vs 73.0%), Terminal-Bench-Science (77.1% vs 68.0%) and AutomationBench (52.4% vs 41.1%). OpenAI suggests Sol for day-to-day coding and Astra for the hardest professional, research and 3D work.

Versus Claude Opus 5.5 / Sonnet 5.5

Opus 5.5 costs $4 / $20 and scores 59.1 on the Artificial Analysis Intelligence Index at $10.04 per task; GPT-6.2 Sol scores 57.3 at $1.01, about a tenth of the cost. On OpenAI’s Agents’ Last Exam Sol’s 73.0% beats Opus 5.5’s 71.6%, while on FrontierCode 2.0 its 43.8% trails Opus 5.5’s 45.5%. Sonnet 5.5 has the same $2 / $10 price. Pick Claude for top scores and long coding sessions, and Sol for high-volume agent work on a budget.

Setting effort

OpenAI suggests starting at medium and going higher only when the quality gain is worth it. Artificial Analysis measures medium at 52.7 on its index, $0.41 per task and about 2.8 seconds to the first answer token; xhigh at 55.7, $0.65 and about 23 seconds; and max at 57.3, $1.01 and about 62 seconds. ChatGPT Work defaults consumer plans to Light effort. Reasoning tokens bill as output, so don’t default to max for interactive use.

OpenAI latest model guide

Third-party evaluations and first-day reception

  • Artificial Analysis: It scores 57 on the Intelligence Index at max effort, 2 points above GPT-6.1 Sol and level with GPT-6 Astra at about 15% of its cost per task. Running the whole index cost $1,108, or $1.01 per task. Its AA-Omniscience hallucination rate is 55%. Its Coding Agent Index is 3 points below GPT-6.1 Sol, mostly from a 14-point drop on SWE-Atlas-QnA. Every Wokey meter is 9% of the official rate, so the same max-effort task costs about $0.091.
  • Vals AI: At max effort it scores 66.06% on the Vals Index, 9th of 41 models, at $3.18 per test, under a third of GPT-6.2 Astra’s $10.68. It ranks third on Terminal-Bench-Science (58.57%) and scores 60.61% on Terminal-Bench 4.0.
  • VentureBeat (first-day reception): Most first-day criticism targets Astra, but tester Matt Shumer’s “essentially unusable” covered both Astra and Sol, and on X he rated both below Claude Opus 5.5. VentureBeat notes these are a few testers on launch day. On the other side, on X, Tibo of OpenAI’s Codex team calls 6.2 Sol his go-to model for most day-to-day work, Sam Altman singled out Sol with Ultrafast, and tester Max Weinbach called Sol on Ultrafast probably the best model experience to date.

Migrating from GPT-6.1 Sol and using it in Codex

These changes come from OpenAI’s model page, latest model guide and Codex docs. The price is unchanged and most requests only need model set to gpt-6.2-sol, but check the items below first.

  • GPT-6.1 Sol already rejected none and minimal, and 6.2 Sol rejects none too. Move none or minimal to low, and start new integrations at medium.
  • Remove temperature, top_p and top_logprobs as OpenAI requires; Chat Completions also drops logprobs, and Responses removes message.output_text.logprobs from include.
  • Tool calling requires Responses. Through Wokey, Chat Completions requests are converted to Responses before forwarding, so Chat requests with tools also work.
  • From GPT-6.2 on, the API drops reasoning items when it detects mid-conversation edits to the history (prefix edits), a model downgrade, or a change of end user or organization, which can lower answer quality and add tokens. Keep the prefix unchanged and append new instructions at the end; that also improves cache hits.
  • To raise or lower effort mid-conversation, append a configuration_update item between replies instead of changing the request-level reasoning.effort; set cache retention with prompt_cache_options.ttl ("30m", or "12h" in Responses only).
  • The system card shows it is more likely than GPT-6.1 Sol, in two tests, to omit obstacles that blocked its work, such as a missing file or unavailable search tool, so ask automated runs to report blockers explicitly.
  • Above 272K input tokens OpenAI bills the whole request at 2× input and cache and 1.5× output. Wokey has no such tier: the full 1M context is billed at one rate.
  • In Codex, GPT-6.2 Sol’s default context is 500K; set model_context_window in config.toml to go up to about 1M. To use it through Wokey, add the provider below and put your Wokey API key in the WOKEY_API_KEY environment variable.

Using GPT-6.2 Sol in Codex through Wokey (~/.codex/config.toml)

# ~/.codex/config.toml
model = "gpt-6.2-sol"
model_provider = "wokey"
model_reasoning_effort = "medium"

[model_providers.wokey]
name = "Wokey"
base_url = "https://api.wokey.ai"
env_key = "WOKEY_API_KEY"
wire_api = "responses"
Model ID
gpt-6.2-sol
Vendors
openai
Supply status
Currently callable

Catalog pricing snapshot

Prices are per 1M tokens; request-time pricing controls settlement.

MeterWokeyOfficial referenceSavings
Input$0.18$291%
Output$0.9$1091%
Cache read$0.009$0.191%
Cache write$0.225$2.591%

Official price source: https://developers.openai.com/api/docs/models/gpt-6.2-sol

Model capabilities

Context
1,050,000 tokens
Max output
128,000 tokens
Streaming
Supported
Tools
Supported
Vision
Supported
Supported API forms
/v1/chat/completions · /v1/responses

Send a request

/v1/chat/completions

curl https://api.wokey.ai/v1/chat/completions \
  -H "Authorization: Bearer $WOKEY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6.2-sol","messages":[{"role":"user","content":"Hello"}]}'

/v1/responses

curl https://api.wokey.ai/v1/responses \
  -H "Authorization: Bearer $WOKEY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6.2-sol","input":"Hello"}'

Use GPT-6.2 Sol in your tools

The gateway is https://api.wokey.ai (OpenAI-compatible clients usually take https://api.wokey.ai/v1), and the model ID is gpt-6.2-sol. Open the guide for the client you use.

Frequently asked questions

How is GPT-6.2 Sol different from GPT-6.1 Sol?

Same price ($2 / $10, cached input $0.10), more capable. It leads on eight of OpenAI’s nine comparison charts and gains 5.3 points on GDPval v2; the Artificial Analysis Intelligence Index rises from 55 to 57, and the knowledge cutoff moves to 2026-06-01. In Codex, OpenAI estimates 17–180 local messages per 5 hours on Plus (6.1 Sol: 15–160). AA’s Coding Agent Index is 3 points lower, though. Neither accepts none effort, so migrating usually only needs a new model ID.

Should I use GPT-6.2 Sol or GPT-6.2 Astra?

Start with Sol for everyday work. Astra costs four times as much ($8 / $40 vs $2 / $10). On OpenAI’s FrontierCode 2.0 at max effort Sol scores 43.8% and Astra 44.2%, with Sol at about a fifth of the cost per task; the Artificial Analysis Intelligence Index has Astra at 62 and Sol at 57. Astra pulls clearly ahead on professional documents, research, 3D and the hardest long-running tasks, and OpenAI suggests starting day-to-day coding with Sol.

Is GPT-6.2 Sol better than Claude Opus 5.5 or Sonnet 5.5?

On the Artificial Analysis Intelligence Index Opus 5.5 scores 59.1 and GPT-6.2 Sol 57.3, at $10.04 vs $1.01 per task; Opus 5.5 costs $4 / $20, and Sonnet 5.5 has Sol’s $2 / $10 price. On OpenAI’s Agents’ Last Exam Sol’s 73.0% beats Opus 5.5’s 71.6%, while on FrontierCode 2.0 its 43.8% trails Opus 5.5’s 45.5%. Most first-day testers still preferred Claude for long coding sessions; Sol wins on cost per task.

What is Ultrafast on GPT-6.2 Sol, and what does it cost?

Ultrafast is OpenAI’s fastest service tier; on GPT-6.2 Sol it generates up to 4.5× faster than Standard. In the API it costs 3× the Standard rate through 2027-02-28 ($6 / $30) and 6× after, with US data residency only. In ChatGPT and Codex, Pro $200 gets Ultrafast on Sol for the first time and Pro $500 gets it on both models, using included usage at 6×. Fast mode costs 2×.

How much does GPT-6.2 Sol cost, and how is context over 272K billed?

The official price is $2 input, $0.10 cached input and $10 output per 1M tokens; cache writes cost $2.50 for 30-minute retention and $5 for 12 hours. When input exceeds 272K tokens the whole request is billed at $4 / $0.20 / $15. Batch is 50% off ($1 / $5) and Fast mode 2×. Wokey has no long-context tier: the full 1M context is billed at one rate, $0.18 input / $0.90 output / $0.009 cached input per 1M tokens.

Why can’t I find GPT-6.2 Sol in ChatGPT? Is it free?

GPT-6.2 Sol is only in ChatGPT Work and Codex, not the regular Chat view; open Work, where new chats on consumer plans default to Sol at Light effort. Codex docs list Plus, Pro, Business, Enterprise and Edu, with Enterprise and Edu off until an admin enables them; Free and Go are not listed. Without a subscription you can call gpt-6.2-sol in the API and pay per token; through Wokey that is $0.18 input / $0.90 output per 1M tokens.

How far do Codex limits go with GPT-6.2 Sol?

OpenAI estimates 17–180 local messages per 5 hours on Plus, slightly more than GPT-6.1 Sol’s 15–160 and about three times GPT-6.2 Astra’s 6–50. Pro currently has no 5-hour limit, though weekly limits may apply. Codex credits match 6.1 Sol at 50 input / 2.5 cached / 250 output per 1M tokens. Fast uses 2× included usage (down from 2.5×) and Ultrafast 6× (down from 8×). To avoid the 5-hour window entirely, pay per token through the API: through Wokey, $0.18 input / $0.90 output per 1M tokens, charged to your balance, with no 5-hour or weekly limits.

Which reasoning effort should I use with GPT-6.2 Sol? Does it support none?

No, the API rejects none. OpenAI suggests starting at medium and moving none or minimal to low; the launch charts and Codex offer low, medium, high, xhigh and max. Artificial Analysis measures medium at 52.7 on its index with about 2.8 seconds to the first answer token, xhigh at 55.7 and about 23 seconds, and max at 57.3 and about 62 seconds. Use medium for interactive work and consider xhigh for long tasks.

Is there a GPT-6.2 Luna or GPT-6.2 Terra?

Not as of the check date. OpenAI released only GPT-6.2 Sol and GPT-6.2 Astra and has announced no 6.2 Luna or Terra. For a cheaper tier, use GPT-6.1 Terra ($1 / $5) or GPT-6 Luna ($0.10 / $0.50). Luna, Terra, Sol and Astra are tiers from lowest to highest; 6, 6.1 and 6.2 are generations.

How do I use GPT-6.2 Sol in Codex?

ChatGPT Plus, Pro, Business, Enterprise and Edu users can pick it with /model in the Codex app and CLI or run codex --model gpt-6.2-sol, with usage counted against the plan; the default Codex context is 500K, which model_context_window raises. To pay per token, set model to gpt-6.2-sol in ~/.codex/config.toml and add a provider with base_url https://api.wokey.ai, wire_api responses and env_key WOKEY_API_KEY, at $0.18 input / $0.90 output per 1M tokens (official $2 / $10).

How much does the GPT-6.2 Sol API cost, and how much does it save versus the official API?

Wokey lists input at $0.18 per 1M tokens and output at $0.9 per 1M tokens; the official references are $2 and $10. That is 91% lower for input and 91% lower for output. Request-time pricing controls settlement, and the vendor pricing source is linked on this page.

What is the GPT-6.2 Sol API model ID?

The model ID for GPT-6.2 Sol on Wokey is gpt-6.2-sol. Use it exactly as written in the request body's model field or in your client's model setting (for example ANTHROPIC_MODEL in Claude Code, or model in the Codex CLI config.toml); GET https://api.wokey.ai/v1/models also lists it.

Which API forms does GPT-6.2 Sol support, and how do I call it through Wokey?

GPT-6.2 Sol supports /v1/chat/completions (Chat Completions requests) and /v1/responses (Responses requests). Set the API base URL to https://api.wokey.ai, authenticate with a Wokey API key, and use the canonical model ID gpt-6.2-sol; you can start with /v1/responses. The Send a request section includes a curl example for every supported API form.

How do I use GPT-6.2 Sol in Codex CLI?

Codex CLI calls GPT-6.2 Sol over the Responses protocol. Add a model_provider to ~/.codex/config.toml with base_url = "https://api.wokey.ai", wire_api = "responses", and env_key = "OPENAI_API_KEY"; set model to gpt-6.2-sol and put your Wokey API key in OPENAI_API_KEY. The Codex CLI guide has the full config.

What are the context window and maximum output for GPT-6.2 Sol?

The catalog context window is 1,050,000 tokens and the maximum output is 128,000 tokens. These are separate limits, and a request must also satisfy the upstream constraints in effect when it is processed.

How are cache reads and cache writes priced for GPT-6.2 Sol?

Cache read: Wokey lists $0.009 per 1M tokens and the official reference is $0.1 per 1M tokens. Cache write: Wokey lists $0.225 per 1M tokens and the official reference is $2.5 per 1M tokens.

How does calling GPT-6.2 Sol through Wokey differ from OpenRouter?

Wokey lists GPT-6.2 Sol at $0.18 input / $0.9 output per 1M tokens, against an official reference of $2 / $10. OpenRouter states it adds no markup on inference and charges its fee when you buy credits (5.5% by card). Wokey carries a curated set of models and has no per-provider routing parameters, so OpenRouter fits better if you need a wider catalog. The OpenRouter alternative page has the full comparison.

How do GPT-6.2 Sol and GPT-6 Astra differ in price and context?

In the published price comparison, GPT-6.2 Sol lists input/output at $0.18 / $0.9 with a 1,050,000-token context window; GPT-6 Astra lists $0.55 / $2.75 with a 1,050,000-token context window. This is a factual price-and-capacity comparison, not a quality ranking; also compare their supported API forms and capabilities.

Related models and documentation

Get API key · gpt-6.2-sol