---
collection: frontier-ai-labs
collection_name: Frontier AI Labs
week_start: 2026-09-21
title: Two frontier labs ship a generation at once: GPT-6 and Claude Opus 5.5
release_count: 39
generated: 2026-09-28
canonical: https://releases.sh/collections/frontier-ai-labs/digest/2026-09-21
---

# Two frontier labs ship a generation at once: GPT-6 and Claude Opus 5.5

OpenAI introduced GPT-6 Sol and Luna with aggressive token pricing and a prompt-caching overhaul, while Anthropic launched Claude Opus 5.5 as the first model of a new Claude 5.5 family — and made it the default Opus across Claude Code and the Claude API.

### Two new model families land in the same week

OpenAI released GPT-6 Sol and GPT-6 Luna as [reasoning models that accept text and image inputs](https://developers.openai.com/api/docs/changelog#gpt-6-sol-and-gpt-6-luna-released), priced far below their GPT-5.6 predecessors: Sol at $2 input / $10 output per million tokens, Luna at $0.10 / $0.50 for high-volume work. The launch is [positioned as frontier intelligence for everyday work](https://openai.com/index/introducing-gpt-6-sol-and-luna), and the same pairing [rolls into Codex and ChatGPT Work](https://developers.openai.com/codex/changelog/#codex-2026-09-22-gpt-6-sol-luna) — Sol for complex agentic coding, Luna for focused bulk tasks — reaching Plus, Pro, Business, Enterprise, and Edu, with Enterprise admins required to enable them. Prompt caching got its own treatment alongside the launch, with [higher cache hit rates, explicit breakpoints, and new diagnostics](https://openai.com/index/better-prompt-caching-for-gpt-6) to cut latency and cost.

Anthropic answered with [Claude Opus 5.5, the first model in the new Claude 5.5 family](https://support.claude.com/en/articles/12138966-release-notes#claude-opus-5-5-launch), performing at Fable 5.1's level on most work while costing 40% less to run than Opus 5. Anthropic frames it as [matching Fable 5.1 intelligence at higher speed](https://www.youtube.com/watch?v=1f13Bl1sYkw), writing more clearly for long sessions, with raised usage limits on Pro, Max, and Team. On the platform side it ships with a [1M token context window by default, always-on adaptive thinking, and $4/$20 per MTok](https://platform.claude.com/docs/en/release-notes/overview#claude-opus-5-5-1m-context-adaptive-thinking-fast-mode) across the Claude API, Bedrock, Google Cloud, and Microsoft Foundry — a release carrying real breaking changes, since thinking can no longer be disabled and some `tool_choice` values now return 400 errors.

### Codex becomes a cross-platform surface

Codex's week was mostly about catching up to the new models. The CLI [added GPT-6 Sol and Luna with Amazon Bedrock support and migration prompts](https://developers.openai.com/codex/changelog/#github-release-396234532), turned fullscreen transcripts on by default, and fixed proxy routing, network restrictions across redirects, and tmux mouse handling. Before that, the models became [selectable from the model picker, with the rate-limit switch prompt now recommending Luna](https://developers.openai.com/codex/changelog/#github-release-394258789). Codex also arrived on iOS as [ChatGPT for iOS with a redesigned home screen and an iPad split view](https://developers.openai.com/codex/changelog/#codex-2026-09-23-mobile) that keeps the task list beside the open task, plus nested Git repository browsing and side chats that carry over attachments and review comments.

### SDKs race to expose the new models

The official SDKs moved in lockstep. Python [added GPT-6 Sol and Luna model identifiers](https://github.com/openai/openai-python/releases/tag/v3.18.0), then [GCP external storage and the GPT-Rosalind research model](https://github.com/openai/openai-python/releases/tag/v3.19.0) alongside fixes that had clients retrying only replayable content and async methods using `asyncio.get_running_loop`; earlier in the week it [added external storage configuration and safety case retrieval](https://github.com/openai/openai-python/releases/tag/v3.17.0). Two patches followed: one [preserving single-pass chat tool iterables and merging HTTP headers case-insensitively](https://github.com/openai/openai-python/releases/tag/v3.19.1), another [preserving single files for fallback extraction](https://github.com/openai/openai-python/releases/tag/v3.19.2). Go [added the same Sol and Luna identifiers](https://github.com/openai/openai-go/releases/tag/v3.65.0), then [GPT-Rosalind and GCP external storage](https://github.com/openai/openai-go/releases/tag/v3.66.0), after fixing [legacy model choices no longer being preserved](https://github.com/openai/openai-go/releases/tag/v3.64.3). Node [added Sol and Luna identifiers](https://github.com/openai/openai-node/releases/tag/v7.22.0), [GCP external storage and GPT-Rosalind](https://github.com/openai/openai-node/releases/tag/v7.23.0), and [session environment reset events](https://github.com/openai/openai-node/releases/tag/v7.21.0).

### Claude Code rides the new default

Claude Code [made Opus 5.5 the default Opus model with 1M context and $4/$20 pricing](https://code.claude.com/docs/en/changelog#2-1-280), and tightened write approval so a write through a symlinked path is judged by where it lands — no more auto-approved writes landing outside the tree. It also [hardened resume and proxy stream handling](https://code.claude.com/docs/en/changelog#2-1-281), stopping `rm -rf "$(pwd)"`-style commands from running unprompted in auto mode and fixing truncated gateway streams being shown as complete. A [maxProseWidth setting](https://code.claude.com/docs/en/changelog#2-1-282) caps prose in wide terminals, with fixes for 400 errors from undecryptable web search history. Admins got [model governance settings and a /doctor prompt-audit command](https://code.claude.com/docs/en/changelog#2-1-283) to block specific models and flag configs written for older ones. Elsewhere, Anthropic's life sciences group [reported that roughly 950 Claude agents identified a novel CRISPR-like enzyme system](https://www.anthropic.com/news/claude-discovers-novel-enzyme-system), the Claude API added [inline tool definitions in mid-conversation messages](https://platform.claude.com/docs/en/release-notes/overview#claude-opus-5-5-1m-context-adaptive-thinking-fast-mode), and ChatGPT Voice [gained plugin support across web, iOS, and Android](https://help.openai.com/en/articles/6825453-chatgpt-release-notes#use-plugins-in-voice-and-get-work-done-by-speaking).

## Releases covered

### Anthropic

- [Anthropic launches Claude Opus 5.5, first model in new Claude 5.5 family](https://support.claude.com/en/articles/12138966-release-notes#claude-opus-5-5-launch)
- [Anthropic launches Claude Opus 5.5, matching Fable 5.1 intelligence at higher speed](https://www.youtube.com/watch?v=1f13Bl1sYkw)
- [Claude Opus 5.5 launches with 1M context and adaptive thinking](https://platform.claude.com/docs/en/release-notes/overview#claude-opus-5-5-1m-context-adaptive-thinking-fast-mode)
- [Claude Code v2.1.280 defaults to Opus 5.5 and fixes symlink write approval](https://code.claude.com/docs/en/changelog#2-1-280)
- [Claude Code 2.1.281 hardens resume, proxy stream, and auto-mode permission handling](https://code.claude.com/docs/en/changelog#2-1-281)
- [Claude Code v2.1.282 adds maxProseWidth and fixes 400 errors from web search history](https://code.claude.com/docs/en/changelog#2-1-282)
- [Claude Code v2.1.283 adds model governance settings and prompt-audit command](https://code.claude.com/docs/en/changelog#2-1-283)
- [Anthropic says Claude discovered novel CRISPR-like enzyme system ART](https://www.anthropic.com/news/claude-discovers-novel-enzyme-system)

### OpenAI

- [OpenAI API releases GPT-6 Sol and GPT-6 Luna reasoning models](https://developers.openai.com/api/docs/changelog#gpt-6-sol-and-gpt-6-luna-released)
- [OpenAI launches GPT-6 Sol and Luna models](https://openai.com/index/introducing-gpt-6-sol-and-luna)
- [Codex adds GPT-6 Sol and Luna models at lower token prices](https://developers.openai.com/codex/changelog/#codex-2026-09-22-gpt-6-sol-luna)
- [OpenAI GPT-6 improves prompt caching with higher hit rates and explicit breakpoints](https://openai.com/index/better-prompt-caching-for-gpt-6)
- [Codex CLI 0.157.0 adds GPT-6 Sol and Luna models](https://developers.openai.com/codex/changelog/#github-release-396234532)
- [Codex CLI 0.156.1 adds GPT-6 Sol and Luna to model picker](https://developers.openai.com/codex/changelog/#github-release-394258789)
- [OpenAI Codex ships ChatGPT for iOS with iPad split view and redesigned home](https://developers.openai.com/codex/changelog/#codex-2026-09-23-mobile)
- [OpenAI Python SDK v3.18.0 adds GPT-6 Sol and Luna model identifiers](https://github.com/openai/openai-python/releases/tag/v3.18.0)
- [OpenAI Python SDK v3.19.0 adds GPT-Rosalind model and GCP external storage](https://github.com/openai/openai-python/releases/tag/v3.19.0)
- [OpenAI Python SDK v3.17.0 adds external storage config and safety case retrieval](https://github.com/openai/openai-python/releases/tag/v3.17.0)
- [OpenAI Python SDK v3.19.1 fixes chat tool iterable and header merging](https://github.com/openai/openai-python/releases/tag/v3.19.1)
- [OpenAI Python SDK v3.19.2 preserves single files for fallback extraction](https://github.com/openai/openai-python/releases/tag/v3.19.2)
- [OpenAI Go SDK v3.65.0 adds GPT-6 Sol and Luna model identifiers](https://github.com/openai/openai-go/releases/tag/v3.65.0)
- [OpenAI Go SDK v3.66.0 adds GPT-Rosalind model and GCP external storage](https://github.com/openai/openai-go/releases/tag/v3.66.0)
- [OpenAI Go SDK v3.64.3 preserves legacy model choices](https://github.com/openai/openai-go/releases/tag/v3.64.3)
- [OpenAI Node SDK v7.22.0 adds GPT-6 Sol and Luna model identifiers](https://github.com/openai/openai-node/releases/tag/v7.22.0)
- [OpenAI Node SDK v7.23.0 adds GCP external storage and GPT-Rosalind model](https://github.com/openai/openai-node/releases/tag/v7.23.0)
- [OpenAI Node SDK v7.21.0 adds session environment reset events](https://github.com/openai/openai-node/releases/tag/v7.21.0)
- [ChatGPT Voice gains plugin support across web, iOS, and Android](https://help.openai.com/en/articles/6825453-chatgpt-release-notes#use-plugins-in-voice-and-get-work-done-by-speaking)