GPT-5.6 (Sol, Terra and Luna)

2026-07-09 · Publisher : OpenAI · Domain : Artificial Intelligence

OpenAI launches the GPT-5.6 family: Sol (flagship model), Terra (balance of cost and intelligence) and Luna (the most economical), with a new max effort mode and a multi-agent ultra mode.

What changes

  • New naming system: the number designates the generation, while Sol, Terra and Luna are durable tiers that evolve at their own pace.
  • Programmatic tool calling in the Responses API, explicit cache controls, persisted reasoning and multi-agent orchestration in beta.
  • Max effort beyond xhigh; ultra mode that coordinates four agents in parallel by default.
  • Strengthened cybersecurity with OpenAI's strictest safeguards; extended access through Trusted Access for Cyber.
  • Price cuts during the model's life: Luna and Terra on July 30, Sol on August 21.

### Specifications - Context: 1,050,000 tokens (272,000 standard), maximum output 128,000 tokens, text and image input, text output. - Knowledge cutoff: February 16, 2026. Effort: none, low, medium (default), high, xhigh, max. - Availability: ChatGPT, Codex and API; the alias gpt-5.6 points to Sol. - Identifiers: gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna. ### Pricing | Model | Input at launch | Output at launch | Current input | Current output | | --- | --- | --- | --- | --- | | Sol | $5.00 | $30.00 | $4.00 (promo until at least November 21, 2026) | $20.00 | | Terra | $2.50 | $15.00 | $2.00 | $12.00 | | Luna | $1.00 | $6.00 | $0.20 | $1.20 | Price per million tokens (USD). Beyond 272,000 tokens: 2x input, 1.5x output. Cache write billed at 1.25x input. ### Benchmarks | Evaluation | Sol | Terra | Luna | GPT-5.5 | | --- | --- | --- | --- | --- | | Terminal-Bench 2.1 | 88.8% | 87.4% | 84.7% | 85.6% | | SWE-Bench Pro | 64.6% | 63.4% | 62.7% | 59.4% | | DeepSWE v1.1 | 72.7% | 69.6% | 67.2% | 67.0% | | OSWorld 2.0 | 62.6% | 50.2% | 45.6% | 47.5% | | BrowseComp | 90.4% | 87.5% | 83.3% | 84.4% | | GPQA Diamond | 94.6% | 92.9% | 92.3% | 93.6% | | ExploitBench | 73.5% | 52.9% | 33.2% | 47.9% | | ARC-AGI-3 | 7.78% | 0.8% | 0.18% | 0.43% | Scores published by the vendor. The post also compares against competing models. ### Key takeaways The family clearly separates use cases: Sol for critical tasks, Terra for everyday use, Luna for volume. Launch prices dropped quickly, so check before any costing.

Links

Tags : LLM, OpenAI, GPT-5, Sol, Terra, Luna, Agent