Comparing Four Major Model Updates From Last Week
iamfakhrealam · x · 2026-07-11
This post summarizes four notable model updates from last week:
- Grok 4.5 (xAI): Dubbed the "value pick," ranking first in agentic tool use, supporting a 500k context, priced at about $2 / $6 per million tokens—roughly half of Opus. Available on OpenRouter.
- GPT-5.6 (OpenAI): The new flagship line, split into three tiers: sol / terra / luna, revealed on July 9. The author cautions that almost all current benchmarks come from OpenAI itself, advising waiting for neutral reviews before migrating.
- GPT-live (OpenAI): ChatGPT's full-duplex voice capability, allowing simultaneous listening and speaking. The focus is on more natural interaction, not a smarter "brain."
- SWE-1.7 (Cognition): Reaches 1000 tok/s on Cerebras, costing about $1.97 per task. Highly aggressive on speed and price, but currently limited to internal Devin use with no API.
The post ends with a more comprehensive comparison chart, asking readers which one they would actually use.
More from Models
- Google says information agents are coming to AI Pro and Ultra this summer — gaganghotra_ · 2026-07-22
- Poolside’s Laguna S 2.1 gets a two-week free run on Nous Portal — NousResearch · 2026-07-22
- Qwen3.8 Max Preview looks substantially better in a side-by-side test with Kimi K3 — curiousily_ · 2026-07-22
- Moonshot’s Kimi K3 reaches #5 on MathArena as the top open model — xeophon · 2026-07-22
- Google launches Gemini 3.5 Flash Cyber for CodeMender, with limited access for governments — GoogleAI · 2026-07-22
- Gemini 3.5 Flash-Lite beats 3.1 Flash-Lite on long-context retrieval in MRCRv2 — Dillonu · 2026-07-22