Skip to content

Commit d2a8f44

Browse files
committed
doc: opencode go
1 parent 1f1f36a commit d2a8f44

2 files changed

Lines changed: 148 additions & 1 deletion

File tree

packages/web/astro.config.mjs

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -224,7 +224,7 @@ export default defineConfig({
224224
"zh-CN": "使用",
225225
"zh-TW": "使用",
226226
},
227-
items: ["tui", "cli", "web", "ide", "zen", "share", "github", "gitlab"],
227+
items: ["go", "tui", "cli", "web", "ide", "zen", "share", "github", "gitlab"],
228228
},
229229

230230
{
Lines changed: 147 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,147 @@
1+
---
2+
title: Go
3+
description: Low cost subscription for open coding models.
4+
---
5+
6+
import config from "../../../config.mjs"
7+
export const console = config.console
8+
export const email = `mailto:${config.email}`
9+
10+
OpenCode Go is a low cost subscription that gives you reliable access to popular open coding models, with the goal of making AI coding accessible to more people.
11+
12+
:::note
13+
OpenCode Go is currently in beta.
14+
:::
15+
16+
Go works like any other provider in OpenCode. You subscribe to OpenCode Go and
17+
get your API key. It's **completely optional** and you don't need to use it to
18+
use OpenCode.
19+
20+
---
21+
22+
## Background
23+
24+
Open models have gotten really good. They now reach performance close to
25+
proprietary models for coding tasks. And because many providers can serve them
26+
competitively, they are usually far cheaper.
27+
28+
However, getting reliable, low latency access to them can be difficult. Providers
29+
vary in quality and availability.
30+
31+
:::tip
32+
We tested a select group of models and providers that work well with OpenCode.
33+
:::
34+
35+
To fix this, we did a couple of things:
36+
37+
1. We tested a select group of open models and talked to their teams about how to
38+
best run them.
39+
2. We then worked with a few providers to make sure these were being served
40+
correctly.
41+
3. Finally, we benchmarked the combination of the model/provider and came up
42+
with a list that we feel good recommending.
43+
44+
OpenCode Go gives you access to these models for **$10/month**.
45+
46+
---
47+
48+
## How it works
49+
50+
OpenCode Go works like any other provider in OpenCode.
51+
52+
1. You sign in to **<a href={console}>OpenCode Zen</a>**, subscribe to Go, and
53+
copy your API key.
54+
2. You run the `/connect` command in the TUI, select `OpenCode Go`, and paste
55+
your API key.
56+
3. Run `/models` in the TUI to see the list of models available through Go.
57+
58+
:::note
59+
Only one member per workspace can subscribe to OpenCode Go.
60+
:::
61+
62+
The current list of models includes:
63+
64+
- **Kimi K2.5**
65+
- **GLM-5**
66+
- **MiniMax M2.5**
67+
68+
The list of models may change as we test and add new ones.
69+
70+
---
71+
72+
## Usage limits
73+
74+
Go includes generous usage limits with three tiers:
75+
76+
- **5 hour limit** — $4 worth of usage
77+
- **Weekly limit** — $10 worth of usage
78+
- **Monthly limit** — $20 worth of usage
79+
80+
To give you an idea of the limit in terms of tokens, $20 roughly gets you:
81+
82+
- 69 million GLM 5 tokens
83+
- 121 million Kimi K2.5 tokens
84+
- 328 million MiniMax M2.5 tokens
85+
86+
You can view your current usage in the **<a href={console}>console</a>**.
87+
88+
:::tip
89+
If you hit a usage limit, you can continue using the free models available.
90+
:::
91+
92+
Usage limits may change as we learn from early usage and feedback.
93+
94+
---
95+
96+
### Pricing
97+
98+
Below are the prices **per 1M tokens**.
99+
100+
| Model | Input | Output | Cached Read |
101+
| ------------ | ----- | ------ | ----------- |
102+
| GLM 5 | $1.00 | $3.20 | $0.20 |
103+
| Kimi K2.5 | $0.60 | $3.00 | $0.10 |
104+
| MiniMax M2.5 | $0.30 | $1.20 | $0.03 |
105+
106+
---
107+
108+
### Use balance after limits
109+
110+
If you also have credits on your Zen balance, you can enable the **Use balance**
111+
option in the console. When enabled, Go will fall back to your Zen balance
112+
after you've reached your usage limits instead of blocking requests.
113+
114+
---
115+
116+
## Endpoints
117+
118+
You can also access Go models through the following API endpoints.
119+
120+
| Model | Model ID | Endpoint | AI SDK Package |
121+
| ------------ | ------------ | ------------------------------------------------ | --------------------------- |
122+
| GLM 5 | glm-5 | `https://opencode.ai/zen/go/v1/chat/completions` | `@ai-sdk/openai-compatible` |
123+
| Kimi K2.5 | kimi-k2.5 | `https://opencode.ai/zen/go/v1/chat/completions` | `@ai-sdk/openai-compatible` |
124+
| MiniMax M2.5 | minimax-m2.5 | `https://opencode.ai/zen/go/v1/messages` | `@ai-sdk/anthropic` |
125+
126+
The [model id](/docs/config/#models) in your OpenCode config
127+
uses the format `opencode-go/<model-id>`. For example, for Kimi K2.5, you would
128+
use `opencode-go/kimi-k2.5` in your config.
129+
130+
---
131+
132+
## Privacy
133+
134+
The plan is designed primarily for international users, with models hosted in the US, EU, and Singapore for stable global access.
135+
136+
<a href={email}>Contact us</a> if you have any questions.
137+
138+
---
139+
140+
## Goals
141+
142+
We created OpenCode Go to:
143+
144+
1. Make AI coding **accessible** to more people with a low cost subscription.
145+
2. Provide **reliable** access to the best open coding models.
146+
3. Curate models that are **tested and benchmarked** for coding agent use.
147+
4. Have **no lock-in** by allowing you to use any other provider with OpenCode as well.

0 commit comments

Comments
 (0)