Skip to content
Model Season every model has a seasonRadarLabs
Radar · Sep 22, 2026

Anthropic and OpenAI cut prices on the same day: Opus 5.5 at $4 and $20, GPT-6 Sol at $2 and $10

20 itemsRSS

News × data

The story published out there and, below it, what real token traffic from the last 7 days says about who is in it. The headline opens the original article; the model name opens its page here.

Pricing ·

Anthropic and OpenAI cut prices on the same day: Opus 5.5 at $4 and $20, GPT-6 Sol at $2 and $10

On Sep 22 Anthropic launched Claude Opus 5.5 at $4 per million input tokens and $20 output, 40% cheaper than Opus 5 on typical workloads, according to the company. The same day OpenAI launched GPT-6 Sol at $2 and $10 and GPT-6 Luna at $0.10 and $0.50, half the price of the 5.6 generation. Cache reads: $0.20 at Anthropic, a 90% discount at OpenAI. OpenAI's top model costs half of Anthropic's, and Luna costs one fortieth of Opus 5.5.

Launch ·

Claude Opus 5.5 arrives 40% cheaper and 30% faster than Opus 5

Anthropic says Opus 5.5 costs 40% less on typical workloads, generates output 30% faster and charges 60% less for cache reads. In the published tests: 66.4% on Terminal-Bench 4.0, 81.8% on OSWorld 2.0 and 67.7% on Humanity's Last Exam with tools. The company cites a 200,000-line audit in 3 hours, against more than 20 on Opus 5. Fast mode costs $8 and $40 and reaches 2.5 times the speed. Identifier claude-opus-5-5, on AWS, Google Cloud and Azure.

Launch ·

OpenAI launches GPT-6 Sol and Luna at half the price of the 5.6 generation

GPT-6 Sol costs $2 per million input tokens and $10 output; GPT-6 Luna, $0.10 and $0.50, both 50% below their 5.6 equivalents, with a 90% discount on cache. OpenAI says Sol makes half as many factual mistakes and that Luna at high effort matches GPT-5.6 Sol at about a hundredth of the cost. On DeepSWE v1.1: 68.8% and 66.6%. Identifiers gpt-6-sol and gpt-6-luna, in ChatGPT and Codex, Luna also for Free users.

Changes in the data

Traffic · Z.ai (GLM)

GLM 5.3 Flash gained 3.5 points of share

From 9.3% to 12.8% between the last two 7-day windows.

What the data says. Z.ai (GLM) has 15.5% of tokens and 6.3% of estimated spend this week.

See GLM 5.3 Flash on Model Season →
Catalog · Xiaomi

MiMo-V2.6-Flash is now in the catalog

Xiaomi, $0.14 input and $0.28 output per 1M tokens, 1M context and optional reasoning. No traffic volume measured yet.

What the data says. Xiaomi has 5.9% of tokens and 0.9% of estimated spend this week. The lab's most used model is MiMo-V2.5, with 5.24% at $0.17 per 1M.

See MiMo-V2.6-Flash on Model Season →
Catalog · Xiaomi

MiMo-V2.6-Pro is now in the catalog

Xiaomi, $0.43 input and $0.87 output per 1M tokens, 1M context and optional reasoning. No traffic volume measured yet.

What the data says. Xiaomi has 5.9% of tokens and 0.9% of estimated spend this week. The lab's most used model is MiMo-V2.5, with 5.24% at $0.17 per 1M, and this one costs 3.1 times more.

See MiMo-V2.6-Pro on Model Season →
Catalog · Xiaomi

MiMo-V2.6-Pro-UltraSpeed is now in the catalog

Xiaomi, $4.35 input and $8.70 output per 1M tokens, 1M context and optional reasoning. No traffic volume measured yet.

What the data says. Xiaomi has 5.9% of tokens and 0.9% of estimated spend this week. The lab's most used model is MiMo-V2.5, with 5.24% at $0.17 per 1M, and this one costs 31 times more.

See MiMo-V2.6-Pro-UltraSpeed on Model Season →
Catalog · Nex AGI

Nex-N2.5-Pro is now in the catalog

Nex AGI, $0.075 input and $0.25 output per 1M tokens, 256k context, optional reasoning and open weights. No traffic volume measured yet.

What the data says. The model has 0.45% of traffic over the last 7 days, ranked 34th.

See Nex-N2.5-Pro on Model Season →

Other changes

  • CatalogNex-N2.5-Mini is now in the catalog. Nex AGI, $0.025 input and $0.10 output per 1M tokens, 256k context, optional reasoning and open weights. No traffic volume measured yet.
  • CatalogGrok 4.7 is now in the catalog. xAI, $1.60 input and $4.80 output per 1M tokens, 488k context and mandatory reasoning. No traffic volume measured yet. See model →
  • DeprecationSeed 1.6 is set to be deprecated on Nov 11, 2026. 50 days left. It is the first notice in the catalog.
  • DeprecationSeed 1.6 Flash is set to be deprecated on Nov 11, 2026. 50 days left. It is the first notice in the catalog.
  • DeprecationSeed-2.0-Code is set to be deprecated on Nov 11, 2026. 50 days left. It is the first notice in the catalog.
  • DeprecationSeedance 1.5 Pro is set to be deprecated on Nov 11, 2026. 50 days left. It is the first notice in the catalog.
  • RemovedClaude Opus 4 was removed from the catalog. Last listed on Sep 21, 2026. Anyone calling this identifier needs to migrate. See model →
  • RemovedKAT-Coder-Pro V2 was removed from the catalog. Last listed on Sep 21, 2026. Anyone calling this identifier needs to migrate.
  • PriceQwen3 VL 30B A3B Instruct got 30% cheaper. From $0.33 to $0.23 per 1M tokens, blended. Input $0.20 → $0.13, output $0.70 → $0.52. New price since Sep 21. See model →
  • PriceQwen3.6 35B A3B got 21% more expensive. From $0.30 to $0.36 per 1M tokens, blended. Input $0.10 → $0.15, output $0.90 → $1.00. New price since Sep 21. See model →
  • AliasGrok Latest now points to Grok 4.7. It used to point to x-ai/grok-4.6. Anyone using the alias now gets a different model without changing a line of code.
  • ContextClaude Sonnet 4 changed its context window. From 1M to 195k tokens. See model →

In the news: summaries written by AI using only what the cited sources published, reviewed before going live. The numbers in the “The data” strip come from our data, traffic from the last 7 days, never from the summary. Catalog snapshot of Sep 22, 2026, compared with the previous ones since Sep 15, 2026. Traffic through Sep 21, 2026. Price is a 75% input and 25% output blend, and only becomes news when the new value holds for two snapshots. Order follows a fixed relevance formula that weighs the type of change and the model’s share.