Skip to main content

Coding-agent subscriptions compared: plans, limits and billing across 32 providers

A coding plan is a 2026 product category with no single comparison behind it: one provider sells requests per five hours, the next sells credits per month, the third sells seats, and the billing periods hide quarterly and yearly discounts. This page puts the subscription plans of the tracked providers that sell subscriptions into two tables, every row generated from the catalogue behind it, every row stamped with the date we verified it and linked to the provider plan page. 3 tracked providers (Chutes, CrofAI, DeepSeek) sell metered usage only, with no subscription to compare, so they have no rows here.

A credit is not a free tier, and neither is a plan

Three different things share one checkout page. A credit is metered usage money: it buys tokens or requests and runs out. A free tier is recurring access without paying: it renews and it rations. A plan is a subscription: it bills every month and it promises a quota. This article covers plans only: what each subscription costs, what quota it promises, and how the billing period changes the price. One-time signup bonuses and free-credit grants belong to the free-credits article, which keeps them; nothing here duplicates that page.

Paid plans, cheapest first

Cells are the monthly prices the providers list, with billing-period variants as separate rows so a yearly discount never hides inside a monthly cell. Limits are the catalogue wording for what each plan promises. A dash is an unpublished price, not a zero price: four providers show their paid tiers only after login, so those cells say nothing instead of repeating an old figure.

Paid coding subscriptions, cheapest first (capture window 2026-10-26 to 2026-10-27). Prices are per month as the provider lists them; billing-period discounts are named in the row.
ProviderPlan$/moLimitsVerified
Google AIGoogle AI Plus$4.492x access to Gemini versus non-AI subscribers; 400 GB or 2 TB storage; More access to Gemini 3.1 Pro and Deep Research2026-10-26
Muse CodeEveryday Usage$5Send 10-50 prompts every 5 hours, including image and video uploads2026-10-27
Qwen Cloud Token PlanToken Plan$6Full-modality model access under one plan; Dedicated Token Plan API key (sk-sp-... format, separate from PAYG key); Token Plan $6/month confirmed on qwencloud.com/benefits/tokenplan (checked 2026-10-27).2026-10-27
StepFunFlash Mini$6.99100 prompts/5h; 400/wk (~1.5K req/5h); Price confirmed on platform.stepfun.ai/step-plan (checked 2026-10-27).2026-10-27
GrokPlus$8Plus plan per x.ai/pricing2026-10-27
SyntheticGo$9Generous limits and reliable access; Allows extra usage at API pricing; 1 Usage Reset/month2026-10-27
StepFunFlash Plus$9.99400 prompts/5h; 1.6K/wk (~6K req/5h); Price confirmed on platform.stepfun.ai/step-plan (checked 2026-10-27).2026-10-27
OpenCodeGo$10$10/month; the Go page publishes per-model estimated requests per 5 hours and monthly usage; Top up credit if needed; cancel any time2026-10-27
OpenDesignGo$10Up to $120 of model credits / month; Design Plan models only; plan models not included; 10 concurrent tasks; Annual option billed $84/year2026-10-27
Mistral Le ChatPro$14.99Full access to Le Chat for all-day coding and long-running tasks; More messages, web searches and faster responses; All-day coding in the CLI, IDE, or on web; $25.5/mo in API credits included; Student price $5.99/mo2026-10-27
Muse CodeHigh Usage$155x more usage than the Everyday Usage plan2026-10-27
VenicePro$18Unlimited requests; Tiers and prices confirmed on venice.ai/pricing (checked 2026-10-27): Free, Pro $18, Pro Plus $68, Max $200.2026-10-27
Z.AI GLMLite$1810,000 credits / week; Billing periods: $18/month, $14.40/month on the quarterly plan (20% off) and $12.60/month on the yearly plan (30% off)2026-10-26
Google AIGoogle AI Pro$18.994x access to Gemini versus non-AI subscribers; 5 TB storage; YouTube Premium Lite included where available; US$10 in monthly Google Cloud credits2026-10-26
Kilo CodeKilo Pass Starter$19Up to $26.60/mo in credits; Annual billing is a separate option: $588/year across the paid tiers2026-10-27
KimiModerato$19Basic usage; Kimi Code billing: $19/month, or $15/month billed annually at $180/year2026-10-26
AnthropicPro$20More usage than Free; exact limits depend on model, features, and task complexity; Monthly or annual billing; Usage resets on a rolling five-hour session window; paid plans add weekly limits; Claude Pro $20/mo, Max 5x $100, Max 20x $200 confirmed on claude.com/pricing (checked 2026-10-27); annual Pro $17/mo equivalent also listed.2026-10-27
OllamaPro$203 cloud models at a time; 50x more than Free; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.2026-10-27
OpenAI CodexChatGPT Plus$20Expanded Codex usage; exact limits vary by model, task, and surface; Codex inclusion re-verified via the OpenAI help article (checked 2026-10-27): included across ChatGPT plans incl. Free and Go; Codex Cloud needs Plus/Pro tiers or higher. Plan prices per chatgpt.com/pricing.2026-10-27
OpenDesignPlus$20Up to $180 of model credits / month; Design Plan up to $160/month + plan models $20/month; 20 concurrent tasks; Annual option billed $168/year2026-10-27
Mistral Le ChatMistral Team$24.99Secure, collaborative workspace for teams building with AI; Up to 30 GB storage per user; Domain name verification and data export2026-10-27
StepFunFlash Pro$291.5K prompts/5h; 6K/wk (~22.5K req/5h); Price confirmed on platform.stepfun.ai/step-plan (checked 2026-10-27).2026-10-27
GrokPremium$30Premium plan per x.ai/pricing2026-10-27
GrokTeams$30$30+/user/month; team plan2026-10-27
SyntheticSubscription$30500 msg/5h; 1 concurrent/model/pack2026-10-27
KimiAllegretto$395x more requests for Kimi Code; Kimi Code billing: $39/month, or $31/month billed annually at $372/year2026-10-26
OpenCodeGo Plus$40Everything in Go; Higher limits and fewer interruptions; Built for larger, demanding projects2026-10-27
Kilo CodeKilo Pass Pro$49Up to $68.60/mo in credits; Annual billing is a separate option: $588/year across the paid tiers2026-10-27
Alibaba Coding PlanCoding Plan Pro$506,000 requests per 5 hours (sliding window); 45,000 requests per week; 90,000 requests per month; Recommended models: qwen3.7-plus (vision), qwen3.6-plus (vision), kimi-k2.5 (vision), glm-5, and MiniMax-M2.5; More models: qwen3.5-plus (vision), qwen3-max-2026-01-23, qwen3-coder-next, qwen3-coder-plus, and glm-4.7; Supports Claude Code, Cline, Codex, Cursor, Qwen Code, OpenClaw, OpenCode, Kilo CLI, and other coding tools2026-10-27
Muse CodePower Usage$5020x more usage than the Everyday Usage plan2026-10-27
VenicePro Plus$687,500 credits per month; 2-month credit banking (unused credits roll forward); Higher image generation limits on Venice Pro models; Tiers and prices confirmed on venice.ai/pricing (checked 2026-10-27): Free, Pro $18, Pro Plus $68, Max $200.2026-10-27
Google AIGoogle AI Ultra (5x)$79.995x Pro access to Gemini; Starting at 20 TB storage; US$40 in monthly Google Cloud credits; YouTube Premium individual included where available2026-10-26
Z.AI GLMPro$806× Lite usage; Billing periods: $80/month, $64/month on the quarterly plan (20% off) and $56/month on the yearly plan (30% off)2026-10-26
SyntheticBig$90Biggest limits for the biggest projects; Higher priority access; 1 Usage Reset/month2026-10-27
KimiAllegro$9915x more requests for Kimi Code; Kimi Code billing: $99/month, or $79/month billed annually at $948/year2026-10-26
AnthropicMax 5x$1005x Pro usage per rolling five-hour session; Higher output limits; Priority access at high traffic times; Claude Pro $20/mo, Max 5x $100, Max 20x $200 confirmed on claude.com/pricing (checked 2026-10-27); annual Pro $17/mo equivalent also listed.2026-10-27
GrokSuper$100Super plan per x.ai/pricing; AI usage billed separately for Cursor-style agents2026-10-27
OllamaMax$10010 cloud models at a time; 5x more than Pro; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.2026-10-27
OpenAI CodexChatGPT Pro (5x tier)$1005x higher usage than Plus; Pro models and Codex; Exact limits vary by model, task, and surface; Codex inclusion re-verified via the OpenAI help article (checked 2026-10-27): included across ChatGPT plans incl. Free and Go; Codex Cloud needs Plus/Pro tiers or higher. Plan prices per chatgpt.com/pricing.2026-10-27
OpenDesignPro$100Up to $480 of model credits / month; Design Plan up to $360/month + plan models $120/month; 30 concurrent tasks; Annual option billed $720/year2026-10-27
Z.AI GLMMax$16814× Lite usage; Billing periods: $168/month, $134.40/month on the quarterly plan (20% off) and $117.60/month on the yearly plan (30% off)2026-10-26
Google AIGoogle AI Ultra (20x)$189.9920x Pro access to Gemini; Starting at 30 TB storage; US$100 in monthly Google Cloud credits; YouTube Premium individual included where available2026-10-26
Kilo CodeKilo Pass Expert$199Up to $278.60/mo in credits; Annual billing is a separate option: $588/year across the paid tiers2026-10-27
KimiVivace$19930x more requests for Kimi Code; Kimi Code billing: $199/month, or $159/month billed annually at $1,908/year2026-10-26
AnthropicMax 20x$20020x Pro usage per rolling five-hour session; Higher output limits; Priority access at high traffic times; Claude Pro $20/mo, Max 5x $100, Max 20x $200 confirmed on claude.com/pricing (checked 2026-10-27); annual Pro $17/mo equivalent also listed.2026-10-27
GrokUltra$200Ultra plan per x.ai/pricing2026-10-27
Muse CodeMaximum Usage$200Best for the most demanding coding use cases; Above Power Usage2026-10-27
OpenAI CodexChatGPT Pro (20x tier)$20020x usage than Plus; Highest Pro usage tier; New sign-ups and upgrades paused from 2026-09-10; existing subscriptions continue to renew; Codex inclusion re-verified via the OpenAI help article (checked 2026-10-27): included across ChatGPT plans incl. Free and Go; Codex Cloud needs Plus/Pro tiers or higher. Plan prices per chatgpt.com/pricing.2026-10-27
OpenDesignMax$200Up to $900 of model credits / month; Design Plan up to $600/month + plan models $300/month; 30 concurrent tasks; Annual option billed $1,176/year2026-10-27
VeniceMax$20022,500 credits per month; 3-month credit banking; Highest image generation limits on Venice Pro models; Tiers and prices confirmed on venice.ai/pricing (checked 2026-10-27): Free, Pro $18, Pro Plus $68, Max $200.2026-10-27
AI RouterGrowth Annual€990EUR 990/month billed yearly; Quotas measured in routed tokens2026-10-27
AI RouterGrowth Monthly€1190EUR 1,190/month billed monthly; Quotas measured in routed tokens2026-10-27
AI RouterScale Annual€2490EUR 2,490/month billed yearly; Quotas measured in routed tokens2026-10-27
AI RouterScale Monthly€2990EUR 2,990/month billed monthly; Quotas measured in routed tokens2026-10-27

Free, custom and login-gated entry points

Pay-as-you-go tiers are metered usage, not subscriptions, so they have no rows in either table even where the provider sells nothing else. What remains below are free plans, custom-priced tiers and paid tiers whose prices publish only after login.

Free, custom and login-gated entry points (2026-10-26 to 2026-10-27). Dashes are unpublished prices, not zero prices.
ProviderPlan$/moLimitsVerified
Agnes AIStarter—1,500 model requests / 5-hour window; Powered by Agnes-3.0-Flash (~100 TPS, 150 off-peak); Compatible with mainstream coding tools; image understanding, image + video models; Starter, Plus and Pro prices display post-login; tiers confirmed live on the subscribe page (checked 2026-10-27)2026-10-27
Agnes AIPlus—7,500 model requests / 5-hour window; Powered by Agnes-3.0-Flash (~100 TPS, 150 off-peak); Compatible with mainstream coding tools; image understanding, image + video models; Starter, Plus and Pro prices display post-login; tiers confirmed live on the subscribe page (checked 2026-10-27)2026-10-27
Agnes AIPro—30,000 model requests / 5-hour window; Powered by Agnes-3.0-Flash (~100 TPS, 150 off-peak); Compatible with mainstream coding tools; image understanding, image + video models; Starter, Plus and Pro prices display post-login; tiers confirmed live on the subscribe page (checked 2026-10-27)2026-10-27
AI RouterPilot—From EUR 2,500 one-off; Scoped validation with your workloads2026-10-27
AI RouterEnterprise—Custom pricing; EUR 50k+ monthly LLM spend fit2026-10-27
BytePlus ModelArkFree$0500K tokens/LLM; 2M vision tokens2026-10-27
BytePlus ModelArkLite—3x the Free tier usage; Coding Plan Lite/Pro prices render client-side and complete post-login at console.byteplus.com; tiers confirmed live on the coding plan page (checked 2026-10-27)2026-10-27
BytePlus ModelArkPro—5x Lite usage; bundles the full coding agent; Coding Plan Lite/Pro prices render client-side and complete post-login at console.byteplus.com; tiers confirmed live on the coding plan page (checked 2026-10-27)2026-10-27
ClineFree$0Free to install and use; Bring-your-own-key from any provider2026-10-27
ClineEnterprise—Custom pricing; SSO, SLA & dedicated support2026-10-27
DeepInfraFree$0Free tier / no upfront monthly subscription; Free and PAYG $0 entry rows; page live with token pricing (checked 2026-10-27).2026-10-27
GitHub ModelsFree$0Free model experiments with GitHub Models rate limits; github.com/marketplace/models returns 404 on 2026-10-27; Free $0 row kept with GitHub Models docs as source.2026-10-27
GrokFree$0Free within generous limits; Real-time web + X search (Limited)2026-10-27
GroqFree$0Base Free Plan limits; see Groq rate-limit table; Free/Developer $0 entry rows; console.groq.com is login-gated (checked 2026-10-27).2026-10-27
GroqDeveloper$0Base Developer Plan limits; see Groq rate-limit table; Free/Developer $0 entry rows; console.groq.com is login-gated (checked 2026-10-27).2026-10-27
Kilo CodeFree$0$0/mo; No hosted credit required2026-10-27
LLM GatewayFree$0Free tier included on LLM Gateway; Free $0 and Enterprise Custom confirmed on llmgateway.io/pricing (checked 2026-10-27).2026-10-27
LLM GatewaySelf-Hosted$0Unlimited; Self-managed; Free $0 and Enterprise Custom confirmed on llmgateway.io/pricing (checked 2026-10-27).2026-10-27
MimoFree$02026-10-27
MimoLite—60M credits; Token Plan Lite/Standard/Pro/Max prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MimoStandard—200M credits; Token Plan Lite/Standard/Pro/Max prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MimoPro—700M credits; Token Plan Lite/Standard/Pro/Max prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MimoMax—1.6B credits; Token Plan Lite/Standard/Pro/Max prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MiniMaxPlus—~1.7B tokens/month of M3 usage; 3–4 concurrent agents; M Plan Plus/Max/Ultra prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MiniMaxMax—~5.1B tokens/month of M3 usage; 4–5 concurrent agents; Hailuo 2.3 video generation: 3 clips / day; M Plan Plus/Max/Ultra prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
MiniMaxUltra—~12.5B tokens/month of M3 usage; 6–7 concurrent agents; Hailuo 2.3 video generation: 5 clips / day; M Plan Plus/Max/Ultra prices display post-login; tiers confirmed live on the token-plan page (checked 2026-10-27)2026-10-27
NVIDIA NIMFree$0~40 RPM; Rolling free endpoint with current featured models (rotates over time); Free/Standard $0 entry rows; build.nvidia.com live (checked 2026-10-27).2026-10-27
NVIDIA NIMStandard$0Pay-per-use after free tier; Free/Standard $0 entry rows; build.nvidia.com live (checked 2026-10-27).2026-10-27
OllamaFree$01 cloud model at a time; Light usage; PAYG $0, Pro $20/mo ($200/yr), Max $100/mo, Team $500/mo confirmed on ollama.com/cloud (checked 2026-10-27); board carries Free/Pro/Max.2026-10-27
OpenDesignFree$0No recurring model-credit grant; Local CLI plus BYOK only; Bring your own Anthropic, OpenAI or Google key to call provider models2026-10-27
ReplicateFree$06 req/min; Free and PAYG $0 entry rows; model pricing page live (checked 2026-10-27).2026-10-27
RunPodServerless$0Pay-per-use GPU inference; Serverless/Cloud $0 entry rows (PAYG-only, no subscription); runpod.io/pricing live (checked 2026-10-27).2026-10-27
RunPodCloud$0Dedicated GPU instances; Serverless/Cloud $0 entry rows (PAYG-only, no subscription); runpod.io/pricing live (checked 2026-10-27).2026-10-27
SambaNovaDeveloper$0No credit card required; sambanova.ai/pricing returns 404 on 2026-10-27; Developer/PAYG $0 entry rows kept with the homepage as source.2026-10-27
ScalewayFree$0Free Generative API quota per model; Free $0 entry row; console.scaleway.com is login-gated (checked 2026-10-27).2026-10-27
StepFunFree$0Limited requests; Basic models; Free entry row; step-plan page live (checked 2026-10-27).2026-10-27
StepFunFlash Max—5K prompts/5h; 20K/wk (~75K req/5h); Flash Max not listed on the step-plan page on 2026-10-27; last-verified $99 figure removed.2026-10-27
SyntheticFree$0Limited requests2026-10-27
VeniceFree$0Limited requests; Standard models; Tiers and prices confirmed on venice.ai/pricing (checked 2026-10-27): Free, Pro $18, Pro Plus $68, Max $200.2026-10-27

How to read a row

On-demand in this article means the walk-up subscription rate with no commitment beyond the month; annual and quarterly rows name their own effective monthly price. Per-seat rows bill every user separately, one-off rows bill once, and credit quotas renew on the schedule the row names. Limits move without warning on credit-metered plans, so treat any credit figure as the provider stated it on the verification date and check the live plan page before committing a team.

How we checked, and what would change our answer

On 2026-10-27 we re-visited all thirty-five provider plan surfaces: static pricing tables and docs directly, JavaScript pages through a rendered reader, and login-gated platforms far enough to confirm the tiers still sell. Rows whose prices render only after login or in client-side scripts (Z.AI, Kimi, Google AI) carry their 2026-10-26 reads; everything else on this page is stamped 2026-10-27. Plan pages move fast in this category: xAI replaced its whole lineup with Plus, Premium, Super, Ultra and Teams; Meta added a Maximum Usage tier; Synthetic added Go and Big; OpenCode added Go Plus; AI Router replaced self-serve Starter and Pro with pilot, Growth, Scale and Enterprise. A drift test fails the site build when the catalogue moves, so this page is regenerated instead of rotting: a repriced plan, a renamed tier or a newly published login-gated price breaks the build until the row is rewritten.

Sources & provenance