A curated timeline of the models that matter: who shipped what, the specs and license you can actually build on, and what it costs. Every entry is sourced from the developer's own release.
19 models tracked ·
verified through 25 Jul 2026 ·
next review by 24 Aug 2026
Latest change
Claude Opus 5 added in release week: $5/$25 pricing, 1M context, May 2026 knowledge cutoff, developer-reported benchmarks from the system card. Claude Opus 4.8 marked superseded but still available. (24 Jul 2026)All changes
Maintained By
Barry Elad
Barry Elad
Founder & Senior Journalist • 765 Articles
Barry Elad is a seasoned journalist and analyst specializing in finance, technology, AI, and founder of SQ Magazine. He explores the world o...
A sourced timeline of model releases, not a benchmark board. Specs and pricing are as published by the developer; “undisclosed” stays undisclosed, and price is a relative tier with the exact figure on the provider’s page via each row’s source. Informational only.
Row represents GPT-5.6 Sol, the frontier variant; the gpt-5.6 alias routes to Sol. Series also includes Terra (mid, 2.50/15) and Luna (cost-optimized, 1/6). All 1.05M context, Feb 2026 cutoff.
Security & safety
System card and preparedness framework; enterprise data controls.
Added in release week; $5/$25 pricing, 1M context, May 2026 knowledge cutoff per the models overview. SWE-bench Verified figure appears in system card section 8.2 prose, not the summary table
Security & safety
System card published. Cyber classifiers allow source-code vulnerability discovery but block binary scanning, penetration testing, and exploit generation.
Base of the Gemini 3 Pro family. Not on the current Google Gemini API pricing page (superseded by Gemini 3.1 Pro at 2/12 and Gemini 3.5); no longer first-party purchasable, so price cleared under the current-pricing test.
Security & safety
Model card and safety evals; enterprise data controls.
Open-weights release Aug 2025 (HF repo xai-org/grok-2 created 2025-08-22). Context 128K from repo config.json (max_position_embeddings 131072, RoPE-scaled from 8192). Sparse MoE (8 experts, 2 active per token); xAI publishes no headline total-parameter count, so params left absent. Text-only causal LM (Grok1ForCausalLM). Non-commercial Grok 2 Community License; weights-only, no first-party API.
Security & safety
Open weights under the Grok 2 Community License (restricted); deployer owns safety tuning.
Meta operates no first-party inference API and points developers to third-party hosts, so access is weights-only. Represents Llama 4 Maverick (1M context, 17B active / 400B total, 128 experts, released 2025-04-05); Scout variant up to 10M context. Text plus image, no audio or video.
Security & safety
Open weights: you own the data boundary and the safety tuning. Scale-cap clause for very large deployments.
Alibaba current Model Studio API lists qwen3.7-max/plus and qwen3.6-flash; the open Apr-2025 variant qwen3-235b-a22b is not on the current first-party pricing page, so price kept free and access set to weights-only (Apache-2.0). Flagship Qwen3-235B-A22B at 128K context; base Qwen3 is text (Qwen3-VL/Omni add vision). Later 2507 updates raised context to 256K (1M on select variants).
Open family (1B/4B/12B/27B); 4B/12B/27B support vision (text plus image). Built on Gemini 2.0 research. ShieldGemma 2 image safety checker built on Gemma 3.
Security & safety
Open weights under Gemma terms; self-host safety is your responsibility.
Command R+ 08-2024 (current version), a refresh of the original Apr 2024 release; 128K context, CC-BY-NC weights plus paid Cohere API. Superseded by Command A (Mar 2025) and Command A Reasoning.
Security & safety
Weights non-commercial (CC-BY-NC); enterprise license required to ship commercially.
First-party API mistral-large-2407 deprecated 2024-11-30 and retired 2025-03-30 per Mistral docs; superseded by Mistral Large 3 (25.12). Weights remain under Mistral Research License (non-commercial); price cleared and access set to weights-only.
Security & safety
EU-based provider; commercial license with enterprise data terms.
No records match the current filters.
Verification ledger
4 most recent of 4 logged updates
Claude Opus 5 added in release week: $5/$25 pricing, 1M context, May 2026 knowledge cutoff, developer-reported benchmarks from the system card. Claude Opus 4.8 marked superseded but still available.24 Jul 2026
Added developer-reported release benchmarks to 17 model records; two models (Claude Fable 5, Command R+) published no standard benchmark table and stay without one.24 Jul 2026
Round-3 audit correction: DeepSeek-V3 reclassified from open weights to community license — its weights ship under the DeepSeek Model License; MIT covers only the code repository.24 Jul 2026
Tracker launched: 18 model records published after a three-round verification against developer sources; 2 records held pending public documentation.24 Jul 2026
How this tracker is maintained
Every model passes the same checks before it appears, and the row keeps pace with each new release.
01
Sourced
Specs, license and pricing come from the developer’s own release: model card, license file, or pricing page. No benchmark screenshots, no third-hand numbers.
02
Dated
Each row carries its release date and when we last confirmed it. Each version bump gets its own new row.
03
Re-checked
Reviewed every 30 days because the field moves fast. License changes, price cuts and deprecations are logged in the ledger.
Why not a leaderboard?
Live benchmarks already exist and shift daily. This is the durable record: what shipped, when, under which license, at what cost. Row details carry the scores each developer published at release; for current head-to-head rankings, use a live board such as LMArena.
How often is it updated?
Every 30 days, and immediately when a major model ships. The ledger logs each addition and reclassification.
Can I cite this?
Yes. Every row links the developer’s primary source. Use “Cite this tracker” for a ready reference.
Informational only. Specs and pricing reflect developer disclosures at publish time and can change with new versions.
