daily brief ·
Kimi K2.6 Price Cut 38%; Codebuff Overtakes Portkey at 169B Tokens
- GPT-5model
- GLM 5.2model
- Kimi K2.6model
- MiniMax M2.7model
- gpt-oss-120bmodel
- Replittool
- Codebufftool
- Firebase Studiotool
- Kilo Codetool
- Lovabletool
- Zencodertool
- DeepSeek V4 Flashmodel
- GLM 5model
- Kimi K2.5model
- Gemma 4 26B A4Bmodel
- Qwen3 Coder 480B A35Bmodel
- Amazon Q Developertool
- Amptool
- Augment Codetool
- Base44tool
- Bolttool
- Claude Codetool
- Codextool
- Codytool
- Cursortool
- Devintool
- Emergenttool
Prices Heavy repricing across gateways today. Kimi K2.6 fell 38% on input ($0.95→$0.589/MTok) and output ($4→$2.48) via models.dev (evt 33932), with a parallel OpenRouter cut to $0.6/$3.41 (evt 33793). GLM-5.2 moved opposite ways by venue: models.dev raised input +15.9% to $1.12 (evt 33909) while OpenRouter cut it -36.1% to $0.7601 (evt 33763). Qwen3-235B-thinking-2507 dropped 23.3% to $0.23/$2.3 (evt 33947); Mistral Small 3.2-24B fell 25% to $0.075/MTok on both models.dev (evt 33924) and OpenRouter (evt 33841); Qwen3-VL-30B fell 13.3% on both venues (evt 33937, evt 33821). GPT-5.6 variants saw steep cuts: Luna -80% to $0.2/$1.2 across opencode (evt 33893), llmgateway (evt 33878), and crossmodel (evt 33863); Terra -20% to $2/$12 (evt 33868, evt 33862). GPT-OSS-120B rose 5.6% input/14.4% output (evt 34023). Minimax-M2.7 fell 12.5%/12.9% (evt 34013). Qwen3-Coder-480B rose 2.8%/5.2% (evt 34025); DeepSeek-V4-Flash edged up 1.4%/1.8% (evt 33991).
Usage Codebuff overtook Portkey AI in OpenRouter coding tokens, 169453936123 vs 159038688906 over 7 days (+6.5% margin) (evt 34246), after passing 50B tokens/day at 60810658459 the prior day (evt 35329).
Plans Replit Core dropped to $20/mo from $25 (evt 34244). Codex, Cursor, and Claude Code restructured tiers (evts 34236, 34238, 34234); Claude Code Pro now lists $200/yr (evt 34234). Devin dropped its $80/mo Teams price (evt 34239); Base44 delisted four paid tiers (evt 34232); Kilo Code shifted its $588/yr Pass from Starter to Pro (evt 34242).
events in this brief
On the tape
- pricemodels.dev ↗
llmgateway/gpt-5.6-luna input price cut $1→$0.2/MTok (-80%) on models.dev; output $6→$1.2
- pricemodels.dev ↗
opencode/gpt-5.6-luna input price cut $1→$0.2/MTok (-80%) on models.dev; output $6→$1.2
- pricemodels.dev ↗
- pricemodels.dev ↗
openrouter/mistralai/mistral-small-3.2-24b-instruct input price cut $0.1→$0.075/MTok (-25%) on models.dev; output $0.3→$0.2
- pricemodels.dev ↗
- pricemodels.dev ↗
openrouter/qwen/qwen3-vl-30b-a3b-instruct input price cut $0.15→$0.13/MTok (-13.3%) on models.dev; output $0.6→$0.52
- pricemodels.dev ↗
openrouter/qwen/qwen3-235b-a22b-thinking-2507 input price cut $0.3→$0.23/MTok (-23.3%) on models.dev; output $3→$2.3
- priceopenrouter.ai ↗
moonshotai/kimi-k2.6 input price cut $0.95→$0.6/MTok (-36.8%) on OpenRouter; output $4→$3.41
- priceopenrouter.ai ↗
qwen/qwen3-vl-30b-a3b-instruct input price cut $0.15→$0.13/MTok (-13.3%) on OpenRouter; output $0.6→$0.52
- priceopenrouter.ai ↗
mistralai/mistral-small-3.2-24b-instruct input price cut $0.1→$0.075/MTok (-25%) on OpenRouter; output $0.3→$0.2
- pricemodels.dev ↗
hyper/minimax-m2.7 input price cut $0.48→$0.42/MTok (-12.5%) on models.dev; output $1.86→$1.62
- pricemodels.dev ↗
crossmodel/openai/gpt-5.6-terra input price cut $2.5→$2/MTok (-20%) on models.dev; output $15→$12
- pricemodels.dev ↗
hyper/gpt-oss-120b input price raised $0.178→$0.188/MTok (+5.6%) on models.dev; output $0.612→$0.7
- pricemodels.dev ↗
crossmodel/openai/gpt-5.6-luna input price cut $1→$0.2/MTok (-80%) on models.dev; output $6→$1.2
- planreplit.com ↗
- pricemodels.dev ↗
llmgateway/gpt-5.6-terra input price cut $2.5→$2/MTok (-20%) on models.dev; output $15→$12
- rankopenrouter.ai ↗
- priceopenrouter.ai ↗
z-ai/glm-5.2 input price cut $1.19→$0.7601/MTok (-36.1%) on OpenRouter; output $3.74→$2.3888
- plankilocode.ai ↗
- planlovable.dev ↗
Lovable plans changed: Free limits changed; Enterprise limits changed
- planzencoder.ai ↗
Zencoder plans changed: Pro limits changed; Enterprise limits changed
- milestone
- pricemodels.dev ↗
- pricemodels.dev ↗
hyper/glm-5 input price cut $0.88→$0.86/MTok (-2.3%) on models.dev; output $2.78→$2.784
- pricemodels.dev ↗
hyper/kimi-k2.5 input price cut $0.534→$0.528/MTok (-1.1%) on models.dev; output $2.82→$2.79
- pricemodels.dev ↗
hyper/gemma-4-26b-a4b-it output price cut $0.424→$0.388/MTok (-8.5%) on models.dev
- pricemodels.dev ↗
- planaws.amazon.com ↗
- planampcode.com ↗
- planbase44.com ↗
- planbolt.new ↗
- planclaude.com ↗
- plancodebuff.com ↗
- plancursor.com ↗
- plandevin.ai ↗
Devin plans changed: Free limits changed; Teams price no longer listed (was $80/mo)
- planemergent.sh ↗
40 detected events · each links its primary source