MODEL VERSIONS ON THIS PAGE HAVE MOVED — CHECKED 8 OCTOBER 2026
Current line, with published API pricing per million tokens:
●
OpenAI — GPT-6 Sol (paid ChatGPT), GPT-6 Luna ($0.10 / $0.50), GPT-6.1 Sol ($2.00 / $10.00), GPT-6 Astra
●
Anthropic — Claude Opus 5.5, Sonnet 5.5 ($2.00 / $10.00), Haiku 5.5 ($0.10 / $0.50 below 100K), Fable 5.1
●
xAI — Grok 4.3 ($1.25 / $2.50), Grok 4.7 ($2.00 / $6.00)
●
Google — Gemini 3.5 Flash-Lite ($0.30 / $2.50), Gemini 3.8 Flash ($0.75 / $3.75 to 31 Dec), Gemini 4 Argon
The comparison below is kept because the reasoning still holds — how these tiers trade price against capability has not changed. The version numbers and prices have.
Current pricing table →
UPDATED JULY 28 — SANCTIONS CHANGE THE RISK CALCULATION
● Performance comparison: Unchanged — K3 SWE Marathon #1, Opus 5 ARC-AGI-3 30.2%
● Price comparison: Unchanged — K3 $3/M input, Opus 5 $5/M input
● NEW — Moonshot API risk: Entity List designation threatened — US enterprises may need license
● NEW — Self-hosted K3 risk: Lower — Modified MIT weights already released cannot be recalled
● Claude Opus 5 risk: Zero regulatory risk — US company, White House framework signatory, FLI C+
● Recommendation change: Freeze new Moonshot API commitments; existing self-hosted weights lower risk
If you are evaluating K3 vs Opus 5 today: The performance and price case for K3 is unchanged. The regulatory risk has materially increased for the Moonshot-hosted API. If you need zero regulatory risk, Opus 5 is the answer. If you are comfortable with self-hosted K3 on Western infrastructure using already-downloaded Modified MIT weights, the risk is lower — but get a legal opinion before expanding production commitments.
Last updated July 28, 2026. Related: Full sanctions analysis → · Claude Opus 5 full review →