OpenCode (Go)
Coding plans · reviewed by Stan K. · Source: data snapshot · data as of 2026-11-11 · scored at Go
$10/mo for a $60 model budget in real dollars — the most honest pricing in the table. The $40 tier is a trap, and October's credential leak is recent history.
#1 of 10 in AI Coding Plans
Tiers — what each dollar buys
| Tier | $/mo | Annual | Quota & notes |
|---|---|---|---|
| Go scored | $10 | — | $-budget per model (e.g. GLM-5.3-Flash $60/mo, GLM-5.3 $15, 32 models) · 5h=20%, week=50%, month=100% |
| Go Plus | $40 | — | Same model budgets, higher limits — community: only 2–3× the $10 plan |
| Go Teams | $20 | — | Per member: Go plus $5/mo above Go on every model |
Agent support: OpenCode + validated CLIs — OpenCode itself plus validated third-party CLIs: Claude Code, Codex, ZCode, Pi, Hermes, jcode, Kilo Code CLI.
PlanScore breakdown — why 74/100
10 plan criteria × 10 points; PlanScore = mean × 10. Low pillars are visible on purpose — that's the point of a score you can audit. Methodology →
What it does well
- ▸Limits denominated in actual dollars, not points: GLM-5.3-Flash $60/mo for a $10 plan
- ▸32 models on the budget list — DeepSeek, GLM, Kimi, MiniMax, GPT, Claude Haiku, Gemini
- ▸Verified by users: $10 runs 3 parallel agents on DeepSeek Flash with room to spare (HN 50012337)
- ▸Cancel any time; top-up credits possible
Where it costs you
- ▸Go Plus $40 delivers only 2–3× the $10 quota — the bad deal in the lineup (HN 50099056)
- ▸Credential leak incident on 19.10.2026 — security history to weigh
- ▸Budgets only work in OpenCode and validated CLIs, not any agent
- ▸Refunds of credit payments come back as credit; no other refund terms published
Hidden traps — the stuff the pricing page doesn't say
"opencode go $40 plan for 2x-3x usage vs their $10 plan is not such a good deal" (HN 50099056). The $10 tier is the product; 4× the price for ~2–3× the quota inverts the value math that makes this row win the table.
A credential leak incident hit OpenCode in October 2026 (HN 50137544). Fixed or not, check the current security posture before pointing it at production repos.
Claude Code, Codex, ZCode, Pi, Hermes, jcode and Kilo Code CLI are validated — but the plan is not a generic API credit you can point anywhere.
Payments made with credit come back as credit and end the period; no general refund policy is published.
Know the walls before you pay: read the traps above, then decide with eyes open.
See OpenCode (Go) plansAffiliate link where available — never affects the score.
- official OpenCode Go docs — budget model, limits, model list (official) verified
- community HN: $10 runs 3 hermes agents on DSF with room to spare (2026-10) verified
- community HN: $40 = only 2–3× the $10 quota — not a good deal (2026-10) verified
- community HN: $40 on DeepSeek flash ≈ 2× the value of Claude Max 20x (2026-10) verified
- community HN: OpenCode credential leak incident (2026-10-18) verified
Data as of 2026-11-11. Prices and limits change monthly — report anything stale. How we verify: methodology.
Closest alternatives
The only plan that publishes its credit math — 12,000 credits/5h on Pro, half-price off-peak, works
The plan everyone already has access to: $10 buys $15 of multi-model credits (Claude, GPT, Gemini, G
Claude Pro / Max / Ultra — the model everyone else gets compared to, behind the most hated 5-hour +
OpenCode (Go) plan review
In-depth analysis · last updated 2026-11-11
Tiers and discounts
Go $10/mo, Go Plus $40/mo, Go Teams $20/member/mo (+$5/mo above Go on every model). No annual discount — none needed at these prices. The budget model is the whole point: each model gets a monthly dollar limit (GLM-5.3-Flash $60, GLM-5.3 $15, Kimi K3.1 Pro $15, MiniMax M3 $60… 32 models), and the walls are percentages of that budget: 20% per 5 hours, 50% weekly, 100% monthly. Scored at Go $10.
Value math
$10 buying a $60 Flash budget is ×6 on API prices — "$60–$80 worth of Deepseek for $5–10" per HN (50096662). On GLM-5.3 (the big model) the budget is $15, so only ×1.5 — the multiple depends on which model you burn. On DeepSeek V4.1 Flash pricing ($0.032/M mixed at 90/8/2), the $60 budget is ~1.9B tokens/mo.
Worst value tier
Go Plus $40: verified community math says it yields only 2–3× the $10 quota (HN 50099056). If $10 is ×6 value, $40 for 2.5× is ×0.6 per dollar — the inversion that keeps this provider's total from being even higher.
Hours before the wall (community)
- "$10 plan... is enough to run 3 hermes agents on DSF 4.1 and leave plenty of left" (HN 50012337)
- "have yet to figure if I can even consume my quota on DeepSeek Flash" (HN 50065884)
- "The OpenCode Go $40 plan with DeepSeek flash is like 2x the value of the Claude max 20x plan" (HN 50156901)
- No verified lockout-style failure reports — the $-budget walls rarely bind before the month does.
Client lock-in
OpenCode + validated CLIs: the budgets work in OpenCode and the validated list (Claude Code, Codex, ZCode, Pi, Hermes, jcode, Kilo Code CLI) — not a generic API credit for any agent.
Versus Anthropic and Z.ai
The direct community comparison is brutal for Claude: "$40 with DeepSeek flash is like 2x the value of the Claude max 20x plan" (HN 50156901) — and that is the weaker OpenCode tier. Against Z.ai, both are the transparent pair of this table (dollars vs published credits); OpenCode wins raw price ($10) and multi-model breadth, Z.ai wins agent breadth (20+) and has no equal for GLM-5.3 volume via the off-peak window. October's credential leak is the entry in the other column.