OpenCode adds a v2 upgrade path
OpenCode 1.19.0 adds an explicit upgrade to v2. The v2 track uses @opencode/cli, with 2.0.26 verified separately; v1 package downloads remain a historical measurement.
/coding-agentsopencode
Anomaly
OpenCode
Open-source coding agent for terminal, desktop, and web workflows. OpenCode 2 uses the @opencode/cli package; the v1 line remains available as opencode-ai.
CLIDesktopWebActiveOpen sourceBYOKMCP Yes
Recent product updates
Reviewed
Selected releases and additions, with links to the original announcements.
Attributed output
GitHub public-attribution count, 2026-06-09
77.9KAttributed outputGitHub public-attribution count, 2026-06-09Commits and PRs on public GitHub provably produced by this agent. Exact if it is visible through its own bot or workflow, an estimate if measured through attribution it declares (these can swing 30 to 40 percent between measurement runs at multi-million scale), and a floor either way.Index rank · all agents
#14 of 40 CLI agents
#17Index rank · all agents#14 of 40 CLI agentsRanked by agent-opened PRs in the trailing 4 complete weeks, matching the index. Agents without a tracked PR series follow, ordered by total attributed output. Other filters or sorts change the list position.Business signal
Monthly developers · 2026-06-07
7.5M monthlyBusiness signalMonthly developers · 2026-06-07$0 software; Go $10/mo
Entry price
Provider tokens or hosted plan usage
$0 software; Go $10/moEntry priceProvider tokens or hosted plan usage2,787,729
npm demand/wk
Adjusted from raw npm channel
2,787,729npm demand/wkAdjusted from raw npm channelWeekly package downloads as an install-demand floor; adjusted when the vendor's main install channel is no longer npm.61,903,124
npm installs to date
Since 2025-05-26
61,903,124npm installs to dateSince 2025-05-26Total npm downloads of this agent's package(s) summed across every tracked week since it was first published. A floor, like the weekly figure. Where a one-week automated/CI burst inflated the total, the baseline-only figure is shown too.From the public record
What the public commits reveal
81 commits · 11 repositories
Search window: – · captured Oct 14, 2026 (UTC)
The same commits can credit more than one agent
1.2%1 / 81 sampled commits
1 of 81 unique sampled commits carry additional agent attribution (Claude). Counting each agent separately can therefore count the same commit more than once. This is shared attribution, not proof of a handoff or which tool did most of the work.
One repository dominates this visible sample
56.8%46 / 81 sampled commits
Ashahell/ReIncarnation accounts for 46 of the 81 distinct attributed commits returned. This makes the sample useful for studying that project’s workflow, but unsuitable as a picture of the agent’s wider adoption.
Specific work you can inspect
jalfern/retrogames: “Zork: the agent exhausts the map honestly instead of deadlocking in a one-exit room (#146)”. Talqo/talqo: “perf(widget): embed from one non-blocking self-contained script tag (#79)”.
This is a sample of publicly attributed work, not total usage or a market-share estimate. Repositories are not customers. Attribution is declared in metadata, not independently verified tool execution; it does not prove exclusive authorship or autonomy. GitHub flagged incomplete search results; only returned, validated records are described and search totals are not used.
Method and audit data
Up to 100 returned search results ordered by commit date within the audit window; attribution trailers validated and duplicate SHAs removed across repositories. A recent convenience sample, not a random or exhaustive sample. Commit sample dates use the UTC committer timestamp, matching the search window. Excluded 0 commits with five or more distinct recognized agent/model credits in one message; broad credit lists are not used as evidence of a workflow.
Returned records dated Oct 9, 2026–Oct 11, 2026.
Download the dated audit evidence (JSON)Activity Over Time
Weekly attributed output on public GitHub. Output visible through the agent's own bot or workflow is exact; output measured through declared attribution is an estimate. Window: 2025-02-03 to 2026-10-12.
npm Installs Over Time
Weekly npm downloads since 2025-05-26, the install-demand trajectory.
Market
Developer output
Public GitHub commits attributed to opencode
opencode declares attribution on the work it produces.
Last full week (ending 2026-10-18): 5,789 signed commits (+2.6% week over week), search estimate.
Snapshot 2026-06-09, confidence Medium, tracked since 2026-06-09. Read the methodology.
- Index rank · all agents
- #17 of 47 · #14 among CLI agents
- Demand share
- 6.5% of tracked npm demand floor
- Business signal
- 7.5M monthly Monthly developers · 2026-06-07
- GitHub share
- 12.2% of tracked stars
- npm channel
- 2.1M raw downloads/week (npm primary)
Best fit
Terminal-first BYOKOpen-source workflowsMulti-provider local control
Direct competitors
- OpenCode reports very large developer reach; treat it as marketing-site adoption rather than paid-user data.
npm channel caveat: npm primaryExpandHide
opencode shows sustained npm growth through spring 2026.
This measures the npm CLI package only; GitHub stars and package downloads are separate signals.
- Demand floor
- 2.8M / wk
- Latest npm week
- 2.1M
- Peak npm week
- 3.7M
- 4-week average
- 2.8M
Trailing 4-week average
Complete Mon-Sun week when cron-backed
Observed public npm range
As of 2026-10-18
Pricing And Economics
ExpandHide
Pricing snapshot: 2026-10-12. Check the vendor for current terms.
Free open-source agent with BYOK; optional Go subscription is $10/month.
- Entry price
- $0 software; Go $10/mo
- Top self-serve price
- $10/mo
- Pricing unit
- Provider tokens or hosted plan usage
- Token transparency
- BYOK/API retail
- Included usage
- BYOK includes no inference. Go includes model-dependent rolling and monthly usage limits.
- Overage path
- Provider-billed API usage, Console credits, or Go top-up credits.
Retail Token Equivalent
BYOK follows provider rates. Go publishes model-dependent usage limits rather than a universal token allocation.
| Plan | Price | Allowance | Unit | Token / Overage Notes |
|---|---|---|---|---|
| OpenCode / BYOK | $0 software | Inference billed separately by the provider. | Provider tokens | |
| Go | $10/mo | Model-dependent rolling and monthly limits; see the live plan page. | Plan usage |
Unit Economics (Measured)
Mean API-retail cost per completed task on the Artificial Analysis coding agent index (358 tasks across 3 suites, pass@1 over 3 runs, snapshot 2026-06-09). These are pay-per-token API costs, not subscription effective costs.
| Harness | AA index | Cost / task | Tokens / task | Time / task |
|---|---|---|---|---|
| Opencode, Opus 4.7 (medium) | 64.6 | $1.82 | 2.2M | 10m |
Break-even vs API retail on the top harness (Opencode, Opus 4.7 (medium), $1.82/task): Go at $10/mo covers about 5 tasks. Assumes the Artificial Analysis task mix.
Summary checked 2026-10-12.
Free open-source agent with BYOK. Optional hosted inference is available through Console pay-as-you-go or the $10/month Go subscription.
- The v2 npm package is @opencode/cli; opencode-ai is the v1 package.
- Go usage limits vary by model. Subscription fees are not a guaranteed token allocation.
- Existing npm download figures below measure only the v1 package and retain their original capture dates.
Product
Providers
AnthropicOpenAIGoogleOpenRouterAWS BedrockAzureLocal providers
Model families
ClaudeGPTGeminiOpenRouter modelsLocal models
Capabilities
Terminal UIRepo editingProvider routingMCPConfigurable agentsLocal workflows
Tradeoffs
- v1 and v2 have separate package and upgrade paths
- Hosted plans and BYOK have different billing