Claude Mythos 5vsGPT-5.6 Luna
Claude Mythos 5 | GPT-5.6 Luna | |
|---|---|---|
| Specifications | ||
Context windowHow much text the model can hold in mind at once — your question, any documents you attach, the conversation so far, and its own reply. Go past it and the earliest part falls out of view. | 1M | — |
| API pricingUSD per 1M tokens · lower wins · base tier | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | — | $0.20 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | — | $1.20 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | — | $0.02 |
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.20Azure |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | — | $1.20Azure |
| Benchmarks | ||
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 88% | 82.5% |
| BenchmarksPublished by one model only | ||
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | — | 40% |
Gray Swan IPI · k = 1Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. Gray Swan's indirect prompt injection benchmark measures how often such an attack succeeds when the attacker gets a single try. Lower is better. | — | 8.3% |
Gray Swan IPI · k = 10Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 10 tries and counts an attack as successful if any of them works. Lower is better. | — | 38.6% |
Gray Swan IPI · k = 15Prompt injection robustness — Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 15 tries and counts an attack as successful if any of them works. Lower is better. | — | 43.9% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 95.5% | — |
CursorBench v3.2Agentic coding — Cursor's own test of harder, real-world coding tasks inside a code editor, on the refreshed v3.2 task set. Scores aren't comparable with v3.1. Higher is better. | — | 61.1% |
Frontier-Bench v0.1Agentic computer work — A hard, ever-evolving set of real computer tasks — coding, system administration, data work, and more — that an AI agent has to complete on its own. Run by the Harbor / Laude Institute team as the successor to Terminal-Bench (v0.1 is the first release of the task set). The score is the share of tasks solved. Higher is better. | — | 14.3% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 88% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 64.5% | — |
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better. | 85% | — |
Arena Elo (Code)Community preference (code) — Like the text arena, but people vote on which AI writes better code. The votes become a chess-style Elo rating on arena.ai. Higher is better. | — | 1517 |
| Overview | ||
| Company | Anthropic | OpenAI |
| Release date | Jun 9 2026 | Jun 26 2026 |
| Access | Proprietary | Proprietary |
Other comparisons
Claude Mythos 5vsGemini 3.7 FlashGPT-5.6 LunavsGemini 3.7 FlashClaude Mythos 5vsMuse GlimmerGPT-5.6 LunavsMuse GlimmerClaude Mythos 5vsGrok 4.6GPT-5.6 LunavsGrok 4.6Claude Mythos 5vsDeepSeek-V4-Pro-0813GPT-5.6 LunavsDeepSeek-V4-Pro-0813Claude Mythos 5vsMistral Medium 3.5GPT-5.6 LunavsMistral Medium 3.5Claude Mythos 5vsKimi K3GPT-5.6 LunavsKimi K3
Frequently asked questions
Claude Mythos 5 leads GPT-5.6 Luna on 1 of the 1 benchmark they both report (Terminal-Bench 2.1). Only GPT-5.6 Luna has a verified first-party API price: $0.20 per million input tokens and $1.20 per million output tokens. No pay-as-you-go API rate is tracked for Claude Mythos 5. Claude Mythos 5 shipped 17 days before GPT-5.6 Luna, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On Terminal-Bench 2.1, Claude Mythos 5 leads at 88% vs GPT-5.6 Luna at 82.5%.
Claude Mythos 5 was released by Anthropic on Jun 9 2026.
GPT-5.6 Luna was released by OpenAI on Jun 26 2026.
Claude Mythos 5 leads on Terminal-Bench 2.1 — Claude Mythos 5 88% vs GPT-5.6 Luna 82.5%.
Only GPT-5.6 Luna has a verified first-party API price: $0.20 per million input tokens and $1.20 per million output tokens. No pay-as-you-go API rate is tracked for Claude Mythos 5. Rates are pay-as-you-go API prices verified on August 18, 2026.