What the coding tools changed

The top-ranked model for Best coding model, Free changed from nvidia/nemotron-3-ultra-550b-a55b:free to stealth/space-bunny-alpha, measured on token volume through OpenRouter.

The top-ranked model for Best model in Roo Code changed from deepseek/deepseek-v4-flash-0731:batch to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Above $5 / M changed from anthropic/claude-opus-5 to moonshotai/kimi-k3, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from poolside/laguna-s-2.1:free to nvidia/nemotron-3-ultra-550b-a55b:free, measured on token volume through OpenRouter.

The top-ranked model for Best model in Cline changed from deepseek/deepseek-v4-flash-0731:batch to z-ai/glm-5.3-flash:batch, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, $1 to $5 / M changed from tencent/hy4-preview to nvidia/nemotron-3-ultra-550b-a55b, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from deepseek/deepseek-v4-flash-0731:free to poolside/laguna-s-2.1:free, measured on token volume through OpenRouter.

The top-ranked model for Best model in Roo Code changed from deepseek/deepseek-v4-flash-0731:free to deepseek/deepseek-v4-flash-0731:batch, measured on token volume through OpenRouter.

The top-ranked model for Best model in Cline changed from deepseek/deepseek-v4-flash-0731:free to deepseek/deepseek-v4-flash-0731:batch, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Above $5 / M changed from moonshotai/kimi-k3:batch to anthropic/claude-opus-5, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, $1 to $5 / M changed from nvidia/nemotron-3-ultra-550b-a55b to tencent/hy4-preview, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from poolside/laguna-s-2.1:free to deepseek/deepseek-v4-flash-0731:free, measured on token volume through OpenRouter.

The top-ranked model for Best model in Roo Code changed from deepseek/deepseek-v4-flash-0731:batch to deepseek/deepseek-v4-flash-0731:free, measured on token volume through OpenRouter.

The top-ranked model for Best model in Cline changed from deepseek/deepseek-v4-flash-0731:batch to deepseek/deepseek-v4-flash-0731:free, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Above $5 / M changed from anthropic/claude-opus-5 to moonshotai/kimi-k3:batch, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, $1 to $5 / M changed from z-ai/glm-5.2:batch to tencent/hy4-preview, measured on token volume through OpenRouter.

The top-ranked model for Best coding model overall changed from deepseek/deepseek-v4-flash-0731 to z-ai/glm-5.3-flash, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from deepseek/deepseek-v4-flash-0731 to z-ai/glm-5.3-flash, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from z-ai/glm-5.2:free to poolside/laguna-s-2.1:free, measured on token volume through OpenRouter.

The top-ranked model for Best model in Aider changed from stealth/ox-alpha to thinkingmachines/inkling:free, measured on token volume through OpenRouter.

Gone: 12.2m tokens; 206m tokens; 313m tokens; 34.5m tokens; 48.1m tokens; 61.7m tokens. New: 1.58m tokens; 24.7m tokens; 25.9m tokens; 316m tokens; 35.6m tokens; 47.7m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. Newly shown: $0.15. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 10.2m tokens; 11.1m tokens; 125m tokens; 17.6m tokens; 3.82m tokens; 30.4m tokens. New: 12.2m tokens; 206m tokens; 30 minutes; 313m tokens; 34.5m tokens; 48.1m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

A Unlimited tier at $40 was on the Devin pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

Gone: 200 tokens; 22.7m tokens; 233m tokens; 35.1m tokens; 50.8m tokens; 566m tokens. New: 10.2m tokens; 11.1m tokens; 125m tokens; 17.6m tokens; 3.82m tokens; 30.4m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: unlimited flex seats; unlimited team members share and. New: 200 users. Plan limits often change without the price changing, which is why this is tracked separately.

A Enterprise tier priced at $40 now appears on the Devin pricing page and did not appear on the previous check.

New: 4.5 prompt; 4.6 prompt; 4.7 prompt; 4.8 prompt; 5 prompt. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 10 minutes; 205m tokens; 211m tokens; 272k tokens; 63.6m tokens. New: 22.7m tokens; 233m tokens; 35.1m tokens; 50.8m tokens; 566m tokens; 8.37m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 164m tokens; 234m tokens; 71.9m tokens. New: 205m tokens; 211m tokens; 63.6m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 214m tokens; 77.8m tokens; 96.5m tokens. New: 164m tokens; 234m tokens; 71.9m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 210m tokens; 24.1m tokens; 78.1m tokens. New: 214m tokens; 77.8m tokens; 96.5m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.18. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 183m tokens; 67.8m tokens; 705m tokens. New: 200 tokens; 210m tokens; 24.1m tokens; 78.1m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The Pro tier on the Subframe pricing page moved from $29 to $20, a 31% decrease. Plan names are matched by position on the page, so check the source before quoting this.

Gone: 24 hour; unlimited ai usage custom fonts; unlimited pages unlimited prototypes; unlimited projects unlimited pages; unlimited prototypes unlimited team; unlimited team members limited ai. New: 3 build; unlimited projects unlimited viewers; unlimited viewers. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1,107 tokens; 166m tokens; 53.8m tokens; 637m tokens. New: 183m tokens; 67.8m tokens; 705m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 145m tokens; 44.5m tokens; 585m tokens. New: 166m tokens; 53.8m tokens; 637m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: unlimited workspaces select core pro. New: unlimited workspaces try now pro. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: unlimited projects to organize chats. Plan limits often change without the price changing, which is why this is tracked separately.

The Pro tier on the Claude Code pricing page moved from $20 to $17, a 15% decrease. Plan names are matched by position on the page, so check the source before quoting this.

New: 500 users. Plan limits often change without the price changing, which is why this is tracked separately.

A Premium tier at $200 was on the Kiro pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Teams tier priced at $200 now appears on the Kiro pricing page and did not appear on the previous check.

Gone: 115m tokens; 32m tokens; 495m tokens; 765m tokens; 803m tokens. New: 145m tokens; 44.5m tokens; 585m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 19.8m tokens; 228m tokens; 392m tokens; 516m tokens; 614m tokens; 90.8m tokens. New: 115m tokens; 32m tokens; 495m tokens; 765m tokens; 803m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the v0 pricing page changed. Newly shown: $90. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 10 pages; 2 prototypes. New: unlimited pages unlimited prototypes. Plan limits often change without the price changing, which is why this is tracked separately.

A Pro tier at $20 was on the Cursor pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Individual tier priced at $20 now appears on the Cursor pricing page and did not appear on the previous check.

Gone: 11.3m tokens; 224m tokens; 257m tokens; 263m tokens; 436m tokens; 48.7m tokens. New: 19.8m tokens; 228m tokens; 392m tokens; 516m tokens; 614m tokens; 90.8m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

A Teams tier priced at $40 now appears on the Cursor pricing page and did not appear on the previous check.

Gone: 119m tokens; 165m tokens; 2.05m tokens; 209m tokens; 211m tokens; 372m tokens. New: 11.3m tokens; 224m tokens; 257m tokens; 263m tokens; 436m tokens; 48.7m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.10. Newly shown: $0.18. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 135m tokens; 142m tokens; 22.6m tokens; 46.9m tokens; 466m tokens; 49.1m tokens. New: 10 minutes; 119m tokens; 165m tokens; 2.05m tokens; 209m tokens; 211m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

A Teams tier at $40 was on the Cursor pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

Gone: unlimited workspaces remove. New: 30 hours; 60 projects; unlimited workspaces select core pro. Plan limits often change without the price changing, which is why this is tracked separately.

A Individual tier at $20 was on the Cursor pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

Gone: 110m tokens; 16.5m tokens; 17 hours; 34.1m tokens; 37.3m tokens; 39.4m tokens. New: 135m tokens; 142m tokens; 22.6m tokens; 272k tokens; 46.9m tokens; 466m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.05. Newly shown: $0.03. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

New: unlimited files and projects team-wide. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the Replit Agent pricing page changed. No longer shown: $17, $95. Newly shown: $18, $90. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 1,000 hours; 750 hours; unlimited usage of your other linked ai. New: 45,000 minutes. Plan limits often change without the price changing, which is why this is tracked separately.

A Pro tier priced at $20 now appears on the Cursor pricing page and did not appear on the previous check.

A Unlimited tier at $20 was on the Amp pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Standard tier at $20 was on the Amp pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Individual tier priced at $20 now appears on the Amp pricing page and did not appear on the previous check.

The Pro tier on the Zed pricing page moved from $20 to $5, a 75% decrease. Plan names are matched by position on the page, so check the source before quoting this.

Gone: 15 hours; 25.2m tokens; 26.4m tokens; 27.7m tokens; 3 hours; 533m tokens. New: 110m tokens; 16.5m tokens; 17 hours; 34.1m tokens; 37.3m tokens; 39.4m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.06. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Plus tier at $200 was on the Amp pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

Gone: 11.7m tokens; 13 hours; 17.8m tokens; 19.2m tokens; 61.1m tokens; 61m tokens. New: 15 hours; 25.2m tokens; 26.4m tokens; 27.7m tokens; 533m tokens; 80.7m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 11 hours; 11.8m tokens; 14.9m tokens; 46.3m tokens; 53.7m tokens; 9.55m tokens. New: 11.7m tokens; 13 hours; 17.8m tokens; 19.2m tokens; 61.1m tokens; 61m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 10 hours; 213k tokens; 33.2m tokens; 36.3m tokens; 5.19m tokens; 5.34m tokens. New: 11 hours; 11.8m tokens; 14.9m tokens; 46.3m tokens; 53.7m tokens; 9.55m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 266m tokens; 331m tokens; 4 hours; 4.1m tokens; 529m tokens; 783m tokens. New: 10 hours; 213k tokens; 3 hours; 33.2m tokens; 36.3m tokens; 5.19m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.01, $0.60. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 1.09m tokens; 1.8m tokens; 10.3m tokens; 178m tokens; 2.8m tokens; 265m tokens. New: 266m tokens; 331m tokens; 4 hours; 4.1m tokens; 529m tokens; 783m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.12, $0.30, $0.72. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: unlimited teams shared libraries and. Plan limits often change without the price changing, which is why this is tracked separately.

A Enterprise tier priced at $20 now appears on the Claude Code pricing page and did not appear on the previous check.

New: 10k requests. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the Firebase Studio pricing page changed. Newly shown: $0.01, $0.06. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 1.23m tokens; 11m tokens; 15 hours; 2.96m tokens; 254m tokens; 264m tokens. New: 1.09m tokens; 1.8m tokens; 10.3m tokens; 178m tokens; 2.8m tokens; 265m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.03, $0.04, $0.21. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: unlimited files and folders for a; unlimited files and folders team-wide; unlimited teams and basic security want. New: unlimited files and projects for a. Plan limits often change without the price changing, which is why this is tracked separately.

The Enterprise tier on the Augment Code pricing page moved from $100 to $20, a 80% decrease. Plan names are matched by position on the page, so check the source before quoting this.

The Team tier on the Augment Code pricing page moved from $100 to $20, a 80% decrease. Plan names are matched by position on the page, so check the source before quoting this.

A Standard tier priced at $20 now appears on the Augment Code pricing page and did not appear on the previous check.

Gone: 1.06m tokens; 10.1m tokens; 16 hours; 195m tokens; 2 hours; 2.45m tokens. New: 1,107 tokens; 1.23m tokens; 11m tokens; 15 hours; 2.96m tokens; 254m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.07. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Recorded default moved from none to Base 1. Wix announced it in its Q2 2026 results, released 4 August 2026: "Base44 becomes first app-creation platform to launch proprietary LLM, Base 1", which Wix says gives it "greater direct control over the largest cost driver in AI-native businesses, compute and inference spend" and takes Base44 from near-zero to roughly 60% non-GAAP gross margin in the second half. Wix does not publish which model serves an individual generation, so read this as the model the product now runs on rather than a per-request guarantee.

Gone: 127m tokens; 14 hours; 167m tokens; 178m tokens; 2.06m tokens; 2.43m tokens. New: 1.06m tokens; 10.1m tokens; 16 hours; 195m tokens; 2.45m tokens; 2.59m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The top-ranked model for Best coding model overall changed from stealth/ox-alpha to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Best model in Cline changed from deepseek/deepseek-v4-flash-0731 to deepseek/deepseek-v4-flash-0731:batch, measured on token volume through OpenRouter.

The top-ranked model for Best model in Roo Code changed from stealth/ox-alpha to deepseek/deepseek-v4-flash-0731:batch, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from stealth/ox-alpha to z-ai/glm-5.2:free, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from z-ai/glm-5.2:free to stealth/ox-alpha, measured on token volume through OpenRouter.

The top-ranked model for Best coding model overall changed from deepseek/deepseek-v4-flash-0731 to stealth/ox-alpha, measured on token volume through OpenRouter.

The top-ranked model for Best model in Roo Code changed from deepseek/deepseek-v4-flash to stealth/ox-alpha, measured on token volume through OpenRouter.

The top-ranked model for Best model in Aider changed from deepseek/deepseek-v4-flash to stealth/ox-alpha, measured on token volume through OpenRouter.

The top-ranked model for Best model in Kilo Code changed from stepfun/step-3.7-flash to poolside/laguna-s-2.1:free, measured on token volume through OpenRouter.

The top-ranked model for Best coding model overall changed from z-ai/glm-5.2 to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Most-used coding model, Free changed from poolside/laguna-s-2.1:free to z-ai/glm-5.2:free, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from z-ai/glm-5.2 to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from deepseek/deepseek-v4-flash-0731 to z-ai/glm-5.2, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Above $5 / M changed from moonshotai/kimi-k3 to anthropic/claude-opus-5, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Free changed from nvidia/nemotron-3-ultra-550b-a55b:free to poolside/laguna-s-2.1:free, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Above $5 / M changed from anthropic/claude-opus-4.8 to moonshotai/kimi-k3, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from tencent/hy3 to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from z-ai/glm-5.2 to tencent/hy3, measured on token volume through OpenRouter.

The top-ranked model for Best model in Cline changed from stepfun/step-3.7-flash to deepseek/deepseek-v4-flash-0731, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from tencent/hy3 to z-ai/glm-5.2, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from stepfun/step-3.7-flash to tencent/hy3, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from z-ai/glm-5.2 to stepfun/step-3.7-flash, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, $1 to $5 / M changed from nvidia/nemotron-3-ultra-550b-a55b to z-ai/glm-5.2, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, $1 to $5 / M changed from z-ai/glm-5.2 to nvidia/nemotron-3-ultra-550b-a55b, measured on token volume through OpenRouter.

The top-ranked model for Best coding model, Under $1 / M changed from stepfun/step-3.7-flash to z-ai/glm-5.2, measured on token volume through OpenRouter.

firebase.studio carries a banner: "We're sunsetting Firebase Studio on March 22, 2027 so you can no longer create new workspaces. Move your code to Google AI Studio or Google Antigravity." The tool still runs until then, which is why this is a sunset rather than a shutdown, but it is a dated end, and starting a project on it now buys under eight months.

continue.dev now reads "Continue has joined Cursor · Continue was acquired by Cursor", and the pricing page it used to publish returns 404. The Apache-2.0 repository is still public and still being pushed to, so the open-source client survives the deal, but Continue is no longer an independent product, and every "Cursor vs Continue" comparison now compares a tool to something its owner bought. No announcement date is published on the page; this is the date it was first observed.

Gone: 1.21m tokens; 1.57m tokens; 122m tokens; 13 hours; 141m tokens; 3 hours. New: 127m tokens; 14 hours; 167m tokens; 178m tokens; 2 hours; 2.06m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1.15m tokens; 1.41m tokens; 11 hours; 125m tokens; 2 hours; 40.6m tokens. New: 1.21m tokens; 1.57m tokens; 122m tokens; 13 hours; 141m tokens; 3 hours. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. Newly shown: $0.06. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Teams tier priced at $60 now appears on the Factory AI pricing page and did not appear on the previous check.

Gone: 15.1m tokens; 3.38m tokens; 357m tokens; 39.9m tokens; 517k tokens; 8 hours. New: 1.15m tokens; 1.41m tokens; 11 hours; 125m tokens; 40.6m tokens; 5.15m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

New: 10 seats. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. Newly shown: $0.01, $0.10, $0.12, $0.21, $0.30, $0.60, $0.72. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Premium tier priced at $39 now appears on the GitHub Copilot pricing page and did not appear on the previous check.

Gone: 305m tokens; 33.4m tokens; 407m tokens; 5 hours; 67.1m tokens; 8.6m tokens. New: 15.1m tokens; 2 hours; 3.38m tokens; 357m tokens; 39.9m tokens; 517k tokens. Plan limits often change without the price changing, which is why this is tracked separately.

New: 2,000 completions; 2.9x+ included usage; 4x+ included usage; unlimited code completion and next edit. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. Newly shown: $0.04, $0.05. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Premium tier at $39 was on the GitHub Copilot pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

Gone: 2,000 completions; 2.9x+ included usage; 4x+ included usage; unlimited code completion and next edit. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the Replit Agent pricing page changed. No longer shown: $25. Newly shown: $17. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

The text "Coding Agent in IDE, CLI and Cloud Kilo has been acquired by Anaconda! Read the full announcement: 🐍 Anaconda Blog Skip to main c" appears on https://kilo.ai and did not appear on the previous check. Read off a marketing page rather than an API, so it is recorded as a signal to check, not as a status change, the status column moves only when it is curated.

Gone: 141m tokens; 15.2m tokens; 24.7m tokens; 267m tokens; 3.12m tokens; 311m tokens. New: 305m tokens; 33.4m tokens; 407m tokens; 5 hours; 67.1m tokens; 8.6m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 112m tokens; 12.5m tokens; 18.7m tokens; 223m tokens; 299m tokens. New: 141m tokens; 15.2m tokens; 24.7m tokens; 267m tokens; 3.12m tokens; 311m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 116m tokens; 125m tokens; 291m tokens; 416m tokens; 9.28m tokens; 9.88m tokens. New: 112m tokens; 12.5m tokens; 18.7m tokens; 223m tokens; 299m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.04. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: 101m tokens; 110m tokens; 2.23m tokens; 246m tokens; 378m tokens; 47 hours. New: 116m tokens; 125m tokens; 291m tokens; 416m tokens; 9.28m tokens; 9.88m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. No longer shown: $0.01. Newly shown: $0.07. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

Gone: $100 monthly credits; 5 collaborators. Plan limits often change without the price changing, which is why this is tracked separately.

A Starter tier at $25 was on the Replit Agent pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Business tier priced at $100 now appears on the Bolt.new pricing page and did not appear on the previous check.

New: 100k project. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1.86m tokens; 212m tokens; 277m tokens; 36 hours; 61.3m tokens; 727m tokens. New: 101m tokens; 110m tokens; 2.23m tokens; 246m tokens; 378m tokens; 47 hours. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1.62m tokens; 158m tokens; 204m tokens; 26 hours; 573m tokens; 69.8m tokens. New: 1.86m tokens; 212m tokens; 277m tokens; 36 hours; 61.3m tokens; 727m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

The set of prices on the OpenRouter pricing page changed. Newly shown: $0.01, $0.03, $0.04. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Plus tier priced at $200 now appears on the Amp pricing page and did not appear on the previous check.

The Unlimited tier on the Amp pricing page moved from $200 to $20, a 90% decrease. Plan names are matched by position on the page, so check the source before quoting this.

New: 1.62m tokens; 158m tokens; 204m tokens; 26 hours; 573m tokens; 69.8m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

New: unlimited public. Plan limits often change without the price changing, which is why this is tracked separately.

A Premium tier at $200 was on the Amp pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Unlimited tier priced at $200 now appears on the Amp pricing page and did not appear on the previous check.

A Team tier at $200 was on the Kiro pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

A Individual tier at $20 was on the Kiro pricing page on the previous check and is not there now. Withdrawn tiers are the change most likely to strand existing users.

The set of prices on the Claude Code pricing page changed. No longer shown: $3, $15. These numbers could not be matched to named plans automatically, so they are reported as a page-level change rather than as a specific plan price.

A Standard tier priced at $20 now appears on the Amp pricing page and did not appear on the previous check.

Gone: unlimited files and projects for a; unlimited files and projects team-wide. New: unlimited files and folders for a; unlimited files and folders team-wide. Plan limits often change without the price changing, which is why this is tracked separately.

New: 2.9x+ included usage; 4x+ included usage; unlimited code completion and next edit; unlimited max plan unlimited copilot; unlimited pro; unlimited with every paid plan. Plan limits often change without the price changing, which is why this is tracked separately.

New: 10 pages; 2 prototypes; unlimited ai usage custom fonts; unlimited projects unlimited pages; unlimited prototypes unlimited team; unlimited team members limited ai. Plan limits often change without the price changing, which is why this is tracked separately.

New: unlimited drafts ui kits and templates; unlimited file storage unlimited cloud; unlimited files and projects for a; unlimited files and projects team-wide; unlimited modes; unlimited sessions. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 1M tokens. New: 1m tokens. Plan limits often change without the price changing, which is why this is tracked separately.

Gone: 10M tokens; 1M tokens; 300K tokens. New: 10m tokens; 1m tokens; 300k tokens; unlimited databases expanded database; unlimited databases pro. Plan limits often change without the price changing, which is why this is tracked separately.

New: 5 apps; unlimited number of apps in-app code. Plan limits often change without the price changing, which is why this is tracked separately.

New: $100 monthly credits; 15 collaborators; 5 collaborators; unlimited workspaces remove. Plan limits often change without the price changing, which is why this is tracked separately.