Step 5 Preview gained 10.6 points of share
From 0.4% to 10.9% between the last two 7-day windows.
What the data says. StepFun has 5.5% of tokens and 7.4% of estimated spend this week.
See Step 5 Preview on Model Season →Step 5 Preview gained 10.6 points of share
Nemotron 3.5 Super VL is now in the catalog
NVIDIA, $0.15 input and $0.50 output per 1M tokens, 1M context, optional reasoning and open weights. No traffic volume measured yet.
What the data says. NVIDIA has 5.2% of tokens and 0.0% of estimated spend this week. The lab's most used model is Nemotron 3 Ultra, with 4.08% at $1.05 per 1M, and this one costs 4.4 times less.
See Nemotron 3.5 Super VL on Model Season →Muse Spark Open is now in the catalog
Meta, $1.25 input and $4.25 output per 1M tokens, 1M context, mandatory reasoning and open weights. No traffic volume measured yet.
What the data says. Meta has 1.6% of tokens and 0.6% of estimated spend this week. The lab's most used model is Muse Spark 1.3 Contributor, with 1.73% at $0.13 per 1M, and this one costs 16 times more.
See Muse Spark Open on Model Season →DeepSeek V4 Pro 0423 got 233% more expensive
From $0.36 to $1.19 per 1M tokens, blended. Input $0.29 → $0.96, output $0.57 → $1.91. New price since Oct 15.
What the data says. The model has 0.39% of traffic over the last 7 days, ranked 34th. DeepSeek has 31.8% of tokens and 7.2% of estimated spend this week. The lab's most used model is DeepSeek V4.1 Flash, with 20.79% at $0.20 per 1M, and this one costs 6.0 times more.
See DeepSeek V4 Pro 0423 on Model Season →GPT-5.6 Sol got 100% more expensive
From $4.00 to $8.00 per 1M tokens, blended. Input $2.00 → $4.00, output $10.00 → $20.00. New price since Oct 15.
What the data says. The model has 0.42% of traffic over the last 7 days, ranked 32nd. OpenAI has 9.1% of tokens and 22.6% of estimated spend this week. The lab's most used model is GPT-6 Luna, with 3.80% at $0.20 per 1M, and this one costs 40 times more.
See GPT-5.6 Sol on Model Season →Other changes
- Pricegpt-oss-120b got 274% more expensive. From $0.070 to $0.263 per 1M tokens, blended. Input $0.037 → $0.150, output $0.17 → $0.60. New price since Oct 15. See model →
- CatalogFLUX 3 Image Klein is now in the catalog. Black Forest Labs, $0.0000 input and $0.0000 output per 1M tokens and 46k context. No traffic volume measured yet.
- CatalogK3 Edit is now in the catalog. Krea, $0.0000 input and $0.0000 output per 1M tokens and 64k context. No traffic volume measured yet.
- Catalogd1-ultra is now in the catalog. Liquid AI, $0.042 input and $0.0000 output per 1M tokens and 32k context. No traffic volume measured yet.
- CatalogMoondream 3 Preview is now in the catalog. Moondream, $0.30 input and $2.50 output per 1M tokens, 31k context, optional reasoning and open weights. No traffic volume measured yet.
- CatalogMoondream 3.1 is now in the catalog. Moondream, $0.30 input and $2.50 output per 1M tokens, 31k context, optional reasoning and open weights. No traffic volume measured yet.
- CatalogSpan-02 is now in the catalog. Respan, $0.010 input and $0.0000 output per 1M tokens and 128k context. No traffic volume measured yet.
- PriceGPT-5.6 Sol Pro got 100% more expensive. From $4.00 to $8.00 per 1M tokens, blended. Input $2.00 → $4.00, output $10.00 → $20.00. New price since Oct 15. See model →
- Trafficjev-1.14 debuted in traffic. First volume recorded on Oct 15, 2026, with 0.08% of traffic over the last 7 days. See model →
- ReasoningGPT-5.1-Codex-Mini now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call.
- Reasoningo1-pro now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call.
- Reasoningo3 Pro now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call.
- PriceGemma 4 31B got 64% more expensive. From $0.15 to $0.25 per 1M tokens, blended. Input $0.090 → $0.200, output $0.34 → $0.40. New price since Oct 15. See model →
- PriceGLM 5 got 67% more expensive. From $0.93 to $1.55 per 1M tokens, blended. Input $0.60 → $1.00, output $1.92 → $3.20. New price since Oct 15. See model →
- Reasoningo1 now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call. See model →
- Reasoningo3 now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call. See model →
- Reasoningo3 Mini now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call. See model →
- Reasoningo4 Mini now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call. See model →
- ReasoningSonar Deep Research now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call.
- ReasoningSonar Reasoning Pro now requires reasoning. Reasoning can no longer be turned off for this model, which increases output tokens on every call.
- RemovedClaude Sonnet 4 was removed from the catalog. Last listed on Oct 15, 2026. Anyone calling this identifier needs to migrate. See model →
- RemovedPareto 26.10 Preview was removed from the catalog. Last listed on Oct 15, 2026. Anyone calling this identifier needs to migrate.
- PriceUniversal-3.5 Pro got 33% cheaper. From $93.75 to $62.50 per 1M tokens, blended. Input $125.00 → $83.33, output $0.0000 → $0.0000. New price since Oct 15.
- PriceQwen3.8 27B got 21% more expensive. From $0.60 to $0.73 per 1M tokens, blended. Input $0.040 → $0.240, output $2.30 → $2.20. New price since Oct 15. See model →
- PriceQwen3.6 35B A3B got 16% more expensive. From $0.31 to $0.36 per 1M tokens, blended. Input $0.10 → $0.15, output $0.95 → $1.00. New price since Oct 15. See model →
- ContextClef Flash changed its context window. From 64k to 24k tokens.
- ContextQwen3 Embedding 8B changed its context window. From 32k to 40k tokens. See model →
Catalog snapshot of Oct 16, 2026, compared with the previous ones since Oct 9, 2026. Traffic through Oct 15, 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.