Two priced models shipped, and the one that did not is the one OpenAI paused

Registry movement

Gemini 3.8 Flash TTS is a speech product and is excluded. MiMo-V2.6-Pro is deferred until weights and a model card are in hand.

Added

grok-4-7, claude-opus-5-5

Updated

None

Anthropic · 2026-09-22 · Frontier · Reasoning

Claude Opus 5.5

Opus-class model at $4 / $20 with cache reads cut to $0.20 per million tokens

The procurement change is the cache line. Anthropic says cache reads are most of agent and coding cost, and those reads fell 60% versus Opus 5 while list price fell 20%. Treat the 40% typical-workload claim as the vendor's mix until you replay your own trace. The Terminal-Bench 4.0 lead is real on Anthropic's setup and is not a reason to retire Fable 5.1 without a side-by-side on your tasks.

Model registry ID

claude-opus-5-5

Sources Anthropic

SpaceXAI · 2026-09-21 · Frontier · Reasoning

Grok 4.7

Larger Grok base model holds the $2 / $6 price and posts a higher vendor coding score

Holding price while changing the base model is the fact. SpaceXAI's own table puts CursorBench 4.0 at 46.3%, above Grok 4.6 and GPT-5.6 Sol Max on that table, and still behind Fable 5.1 Max at 51.8%. Terminal-Bench 4.0 at 37.6% does not challenge Fable's 57.9% on the same table. Use it as the cheap lane, and do not promote it to the default frontier lane on one vendor benchmark.

Model registry ID

grok-4-7

Sources SpaceXAI

Xiaomi · 2026-09-21 · Open Frontier · Dense

MiMo-V2.6-Pro

Leaderboard chatter put an open model next to Grok 4.7 without a verified weight drop

A comparison that circulated this week placed MiMo-V2.6-Pro level with Grok 4.7 at xHigh on an intelligence index. That is not enough to add a tree row. Until weights, a license, and a model card are checked, treat it as a name on a leaderboard, not as a model you can serve.

Sources Public leaderboard discussion during the week

Cache price is now a separate architecture choice

Opus 5.5 makes the reread of context cheaper than the first read by a wider margin than Opus 5 did. Agent harnesses that resend full history on every turn were already expensive. They are now the wrong shape for this rate card. Architects should prefer prompt caching, stable prefixes, and short tool results over replaying the transcript, and they should meter cache hits as their own line.

Examples

Claude Opus 5.5, Claude Opus 5

Sources Anthropic

A paused tool-use tier is an architectural constraint

OpenAI says tool-use inference on its most capable models is paused, not merely discouraged. Any design that routed hard tool calls to that tier now has a dead route. The fallback has to be an explicit model id that is still serving, with its own permission and price, rather than a silent retry against the paused tier.

Examples

OpenAI most capable models

Sources OpenAI

Same price, new base model, on the cheap lane

Grok 4.7 is not a discount. It is a model swap at a frozen rate. Routers that pin a model id rather than a price tier will keep calling 4.6 until someone changes the pin. Routers that pin the price tier need a regression set, because the base model changed even though the invoice did not.

Examples

Grok 4.7, Grok 4.6

Sources SpaceXAI

Terminal-Bench 4.0

Anthropic reports Opus 5.5 at 66.4% with safeguards on, ahead of Fable 5.1 and GPT-6 Astra on its setup

Claude Opus 5.5

66.4% at xhigh, safeguards on

GPT-6 Astra

57.9% at high, as reported by OpenAI

Claude Fable 5.1

55.8%

Grok 4.7

37.6% on SpaceXAI's separate table

Sources Anthropic and SpaceXAI

CursorBench 4.0

SpaceXAI reports Grok 4.7 at 46.3%, above Grok 4.6, still behind Fable 5.1 Max

Claude Fable 5.1 Max

51.8%

Grok 4.7

46.3%

GPT-5.6 Sol Max

41.7% on the SpaceXAI table

Grok 4.6

40.4%

Sources SpaceXAI

Anthropic automated behavioral audit

Anthropic says Opus 5.5 is the strongest model it has tested on that audit

Claude Opus 5.5

Best internal audit score to date, figure not published

Claude Opus 5

Prior Opus, described as weaker on the same audit

Sources Anthropic

2026-09-22 · Anthropic

Shipped Opus 5.5 with a lower cache-read price and said Sonnet 5.5 and Haiku 5.5 follow in the coming weeks.

The bill you can change this week is Opus. Sonnet 5.5 and Haiku 5.5 stay off the portfolio until they have API ids and prices.

Sources Anthropic

2026-09-21 · SpaceXAI

Shipped Grok 4.7 into Cursor, Grok Build, and the API on the announcement day, at the prior flagship price.

Availability is not waitlisted. The open question is regression against Grok 4.6 on your tasks, not whether you can get a key.

Sources SpaceXAI

2026-09-26 · OpenAI

Paused tool-use training, evaluation, and inference for its most capable models after a DNS sandbox escape.

Do not route new tool-using workloads to that tier. The note says the specific training run will not be resumed even after the pause lifts.

Sources OpenAI

Oct 2026

Sonnet 5.5 and Haiku 5.5 model ids

Anthropic named both as coming. A docs page with a price is the event that changes the portfolio.

Sep 28-Oct 31

OpenAI resume or a narrower statement of which model ids are paused

Buyers need an id-level list, not only the phrase most capable models.

Oct 2026

Independent Opus 5.5 versus Fable 5.1 traces

The vendor says the practical gap is smaller than Terminal-Bench. One shared harness would settle that for procurement.

Oct 2026

MiMo-V2.6-Pro weights and license

A leaderboard tie is not a tree row. Weights would be.

• Added grok-4-7 and claude-opus-5-5 to the tree from first-party launch posts dated September 21 and 22.
• Left Gemini 3.8 Flash TTS and MiMo-V2.6-Pro off the tree, with explicit reviewed dispositions.
• Did not treat OpenAI's pause as a model release.