OpenCode Go
Limits, price and how to pay (Nov 2026)
As of 2026-11-09, OpenCode Go costs $10/month, Go Plus $40/month and Go Teams $20 per member per month; you subscribe in OpenCode Console, and the official billing docs list card, Link, Alipay and WeChat Pay among the payment methods. Limits are per-model monthly dollar amounts, not request counts: a 5-hour window can use up to 20% of the monthly limit, a week 50%, a month 100%. For example, DeepSeek V4.1 Flash gets $100 a month on Go and $200 on Go Plus, and Kimi K3 gets $15 and $60. Go and Go Plus are for single-member workspaces and work in OpenCode and other coding agents; if checkout does not work for you or the limits run short, you can connect a per-token API in OpenCode instead (see below).
Updated 2026-11-10
Prices change: on QCode, Claude, GPT and Chinese models are billed per token — live rates on /models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Highlights
Per month
As of 2026-11-09 the official Go page lists Go at $10/month and Go Plus at $40/month, and the docs add Go Teams at $20 per member per month. No $5 first month or coupon is listed. Cancel any time; top up credit if needed.
5-hour window
Each model has a monthly dollar limit: 5 hours 20%, week 50%, month 100%. As of 2026-11-09, on Go DeepSeek V4.1 Flash gets $100 a month and Kimi K3 and GLM-5.3 $15 each; on Go Plus, $200, $60 and $120.
DeepSeek V4.1 Flash per 5h (estimate)
Official estimate from typical usage (about 410 input, 71,300 cached and 310 output tokens per request), as of 2026-11-09; Go Plus: 86,670. Per the official price table, peak-hour rates are twice the off-peak rates, so the same limit buys fewer requests at peak.
Different product
Pay-as-you-go lives in the Console model catalog (often called Zen; opencode.ai/zen now redirects to the Console Models page), while Go is a monthly subscription. With Extra Usage turned on at the Go page, Go falls back to your Console credit once its limits are reached instead of blocking requests.
What Go is
As of 2026-11-09, OpenCode Go is the OpenCode team's (Anomaly) low-cost subscription for a set of tested, popular open coding models (the list also has a few non-open ones such as Claude Haiku 5.5 and GPT 6 Luna). The official docs name three monthly plans: Go at $10/month, Go Plus at $40/month and Go Teams at $20 per member per month. Limits are not request counts but per-model monthly dollar amounts: the tokens you use are priced at that model's rates and deducted from its limit. The request counts on the official page are estimates from typical usage, so judge by the dollar limits and token prices.
How OpenCode Go limits work
Per the Go docs as of 2026-11-09, every model has its own monthly dollar limit, enforced across rolling windows: up to 20% of it in any 5 hours, 50% in a week and 100% in a month; token prices are the same on Go, Go Plus and Go Teams. The limits differ widely: on Go, DeepSeek V4.1 Flash gets $100 a month, MiMo-V2.6-Flash gets $60, and Kimi K3, GLM-5.3, Qwen3.8 Max and Claude Haiku 5.5 get $15 each. In the official request estimates, that is about 110 requests per 5 hours for Kimi K3, about 160 for Qwen3.8 Max and about 43,330 for DeepSeek V4.1 Flash. Go Teams gives every model $5 more monthly limit than Go. OpenCode's explanation: for most models, bulk discounts and reserved GPU capacity make higher limits possible; newer models, or models whose public pricing is already discounted, get lower limits. DeepSeek V4.1 Flash and V4 Pro have peak and off-peak prices; peak hours are 01:00-04:00 and 06:00-10:00 UTC, Monday to Friday. You can track your usage in the console.
Timeline
OpenCode's Go docs updated (page marked Last updated: Oct 6, 2026): Go $10/month, Go Plus $40/month, with monthly dollar limits listed per model.
The OpenCode changelog for v2.2.4 (2026-10-30) says it raised the Go limits for DeepSeek V4.1 Flash: when checked on 2026-10-15 the official docs showed monthly limits of $80 on Go and $160 on Go Plus, and the 2026-11-09 docs show $100 and $200. The same 2026-10-15 check also recorded that DeepSeek V4.1 Flash has peak and off-peak pricing and that the list includes Claude Haiku 5.5.
End date of the current DeepSeek ZDR agreement: the Go docs say it is renewed monthly; checked 2026-11-09, it is valid through 2026-11-30 (it was 2026-10-31 when checked on 2026-10-15). Other changes in the same check: Go Teams ($20 per member per month) has been added; the Go page switch is now called Extra Usage in the docs (the 2026-10-15 docs said Use balance); the billing docs list card, Link, Alipay and WeChat Pay among the payment methods.
Confirmed vs caution
Confirmed
As of 2026-11-09, besides the prices and limit rules above, the official pages and docs confirm: Go and Go Plus are for single-member workspaces such as the Personal workspace created at sign-up, and either can upgrade to Go Teams at any time; Go Teams covers workspaces with several members at $20 per member per month, each with their own limits, and service accounts are not covered; after the limits you can continue on the free models or turn on Extra Usage at the Go page to draw on your Console credit (Go Teams starts with it on); you may qualify for a refund if you ask within 14 days of the charge and have used almost none of that period's allowance, by contacting OpenCode; the DeepSeek ZDR agreement is valid through 2026-11-30; a Console credit purchase carries a processing fee of 4.6% + $0.31 on top.
Caution
Request counts assume typical requests with tens of thousands of cached tokens; longer sessions or more output mean far fewer requests. The model list and limits change (the DeepSeek V4.1 Flash limits were raised on 2026-10-30), and OpenCode says usage limits may change as it learns from early usage and feedback. The $5 first month and coupon codes circulating online could not be found on the official Go page or docs as of 2026-11-09.
How to choose
Go ($10/month)
Fits usage centered on models with higher monthly limits such as DeepSeek V4.1 Flash ($100) and MiMo-V2.6-Flash ($60), with occasional Kimi K3 or Qwen3.8 Max. Remember that any 5 hours can only use 20% of the monthly limit.
Go Plus ($40/month)
If you lean on Kimi K3, GLM-5.3 or Qwen3.8 Max, which get only $15 on Go, Go Plus raises those limits to $60, $120 and $60, and DeepSeek V4.1 Flash to $200. For a team, look at Go Teams ($20 per member per month, with $5 more monthly limit than Go on every model). If you mainly want Claude Opus or Claude Sonnet, they are not on the Go list, so you need a per-token option.
How to use it
① Sign in to OpenCode Console, subscribe to Go, Go Plus or Go Teams, add a payment method under Settings › Billing (the docs list card, Link, Alipay and WeChat Pay among others) with your billing details, and copy your API key. ② Run /connect in the TUI, pick OpenCode Go and paste the key. ③ Run /models to choose a model; in config use opencode-go/<id>, for example opencode-go/kimi-k3. ④ To call Go from another agent (the Go page lists Claude Code, Codex, Hermes and Pi as validated) or your own code, use the endpoints under opencode.ai/inference/go/ (openai/v1/chat/completions, anthropic/v1/messages or openai/v1/responses, depending on the model) and list models at opencode.ai/inference/go/v1/models; your client should use its own user agent and send a stable session ID per conversation in x-opencode-session. ⑤ Track usage in the console.
When Go doesn't fit: a per-token API
When OpenCode Go checkout does not work for you or its monthly limits run short, connect a per-token API in OpenCode and keep working. QCode bills per token, per-model rates are listed on /models, and there is no 4-hour window. One key reaches Claude, GPT and leading Chinese models, for example claude-sonnet-5-5, gpt-6.1-sol, deepseek-v4.1-flash, glm-5.3, kimi-k3 and qwen3.8-flash. In OpenCode, Claude uses the anthropic provider with baseURL https://api.qcode.cc/claude/v1 and GPT uses https://api.qcode.cc/openai/v1; Chinese models need their own @ai-sdk/openai-compatible provider with that same OpenAI baseURL. The full setup is on the OpenCode page at docs.qcode.cc/en/docs/ide/opencode. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; if you already have a Personal Plan, add a booster pack when you need more quota. Details: qcode.cc/pay-as-you-go. Payment options include credit cards, Apple Pay, Google Pay, Alipay, WeChat Pay and crypto.
FAQ
How do OpenCode Go limits work?
By per-model monthly dollar amounts, not request counts. As of 2026-11-09, any 5-hour window can use up to 20% of a model's monthly limit, a week 50% and a month 100%; token prices are the same on Go, Go Plus and Go Teams, and only the monthly limits differ. For example, DeepSeek V4.1 Flash gets $100 a month on Go and $200 on Go Plus; Kimi K3 gets $15 and $60. You can track usage in the console.
How much does OpenCode Go cost?
As of 2026-11-09, Go costs $10/month, Go Plus $40/month and Go Teams $20 per member per month. The plans differ in each model's monthly dollar limit, not in token prices: the official estimate for DeepSeek V4.1 Flash is about 43,330 requests per 5 hours on Go and 86,670 on Go Plus (assuming about 410 input, 71,300 cached and 310 output tokens per request). The real cap is the dollar limit: $100 a month on Go, which by the 20% rule works out to at most $20 per 5 hours. It is priced by peak and off-peak hours; at peak (01:00-04:00 and 06:00-10:00 UTC, Monday to Friday) rates are twice the off-peak rates, so the limit drains faster.
Is there a $5 first month or a coupon?
No. As of 2026-11-09 the official Go page and docs list only Go at $10/month, Go Plus at $40/month and Go Teams at $20 per member per month, with no $5 first month or coupon code. Don't take older $5 first-month claims at face value; the price on the subscribe page is what applies.
What happens when I hit the limit?
As of 2026-11-09 the docs give two paths: keep using the free models, or turn on Extra Usage at the Go page so Go falls back to your Console credit instead of blocking requests (Go Teams starts with it on). If you need more usage, you can also top up credit, move to Go Plus, or connect a per-token API in OpenCode, as described above under when Go doesn't fit.
Refunds and team use?
Refunds: you may qualify if you ask within 14 days of the charge and have used almost none of that period's allowance; contact OpenCode to request one. Teams: Go and Go Plus are for single-member workspaces; a workspace with several members uses Go Teams at $20 per member per month, with their own limits for each member. Both as of 2026-11-09.
How do I buy OpenCode Go, and can I pay with Alipay or WeChat Pay?
You buy it in OpenCode Console, and yes, Alipay and WeChat Pay are among the listed methods. Add a payment method under Settings › Billing and subscribe to Go, Go Plus or Go Teams; as of 2026-11-09 the official billing docs list card, Link, Alipay and WeChat Pay among the supported methods, handled through Stripe, and Go subscriptions use the same methods. The docs do not say which countries or cards are accepted, so the checkout page decides. If Go is not available to you or its limits are too low, connect a per-token API in OpenCode instead: QCode accepts credit cards (Visa / Mastercard), Apple Pay, Google Pay, Alipay, WeChat Pay and crypto, and it is fully self-service. Billed per token; per-model rates are listed on /models. Set it up with the OpenCode page at docs.qcode.cc.
Sources
OpenCode's official Go page (opencode.ai/go, now redirecting to opencode.ai/console/go, including its FAQ), the Go docs (opencode.ai/docs/go/, now redirecting to opencode.ai/docs/console/go/), the billing docs (opencode.ai/docs/console/billing/), the Console models docs (opencode.ai/docs/zen/, now redirecting to opencode.ai/docs/console/models/) and the changelog (opencode.ai/changelog), fetched 2026-11-09; the docs win on numbers.
Compare dollar limits, not request estimates
Prefer per-token billing with one key for claude-sonnet-5-5, gpt-6.1-sol, deepseek-v4.1-flash, kimi-k3, glm-5.3, qwen3.8-flash and more? Try QCode; see /models for each model's price.
Related
Go + DeepSeek
Flash/Pro ids and ZDR.
Go Free
Free tier vs paid Go.
Cline 400 with a custom Anthropic base URL: the fix
4.1.22 started sending the fallbacks option and custom endpoints returned 400; 4.1.23 (2026-10-07) fixes it. Plus the QCode Anthropic-provider setup.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Checked against OpenCode's official Go page, Go docs, billing docs and Console models docs on 2026-11-09; the official pages take precedence. Not affiliated with Anomaly/OpenCode; limits and the model list change. QCode model availability is whatever /models shows.