Repository navigation
Compaction request exceeds context window on high-output models (v2) #42448
Description
Description
A V2 session using a custom OpenAI-compatible model reached about 236k of its 299,964-token context window (79%). Automatic compaction did not run, and /compact failed because the provider reported that the prompt plus requested output exceeded the context limit.
The model advertises a 131,072-token output limit. SessionCompaction.required() caps its reserved output budget at 32,000, but the compaction LLM.request() does not set generation.maxTokens. OpenAI-compatible lowering therefore leaves max_tokens unset, allowing the provider to use 131,072 instead.
Expected: the compaction request should use the same bounded output budget used to calculate the automatic-compaction threshold.
Before /compact, the model also returned finish=stop with no text or tool calls several times, leaving the TUI with no visible response. This may be a separate empty-response issue.
Plugins
No response
OpenCode version
0.0.0-next-17403
Steps to reproduce
- Configure a custom OpenAI-compatible model with context limit 299964 and output limit 131072.
- Grow a V2 session to approximately 236k tokens, displayed as 79% context usage.
- Observe that automatic compaction does not run.
- Run
/compact. - Observe that the provider rejects the request because the prompt plus requested output exceeds the context window.
Screenshot and/or share link
No response
Operating System
No response
Terminal
No response
Activity
bbartels commentedon Aug 14, 2026
@kitlangton Similar to #36253
github-actions commentedon Aug 14, 2026
This issue might be a duplicate of or closely related to existing issues. Please check:
max_output_tokensand leaves oversized sessions unrecoverable — same root area (compaction request construction +maxTokensinteraction with OpenAI-compatible providers), now closed/compactcan exceed model context before summarization on large sessions — similar symptom (compaction failing with context-length error), now closedIf those fixes did not address the specific case of an unset
maxTokenson the compactionLLM.request()allowing the provider to use its full 131k output limit, this is a distinct and still-open issue.bbartels commentedon Aug 17, 2026
@rekram1-node Would you mind having a look? should be a straight-forward fix i believe :)
Zaczero commentedon Aug 18, 2026
The docs already specify the bound this issue asks for —
https://opencode.ai/v2/docs/compaction states that during summary generation "tools disabled and at
most 4096 output tokens" are used. The tools half is implemented (
tools: []); the output half isnot implemented anywhere.
At
39be1595993de57a5a667c86e21ff788a80ebbbe, the compaction request passes nogenerationfield:opencode/packages/core/src/session/compaction.ts
Lines 274 to 282 in 39be159
LLM.requestacceptsgeneration(packages/ai/src/llm.ts:41-51), and OpenAI-compatible loweringreads it at
packages/ai/src/protocols/open-responses.ts:573(
max_output_tokens: generation?.maxTokens), so an unset field leavesmax_output_tokensabsent andthe provider falls back to the model's own limit — 131,072 in this report.
4096appears exactly once inpackages/ai, as an Anthropic-protocol fallback used only when themodel catalog carries no output limit (
anthropic-messages.ts:620). It never applies to theOpenAI-compatible route, so on those providers compaction is unbounded on every model.
So this is a doc/code mismatch as much as a bug: the documented contract is 4096, and passing
generation: { maxTokens: 4096 }on the compaction request would implement it. Worth aligning withSessionCompaction.required(), which already reservesMath.min(limits.output, 32_000)whencomputing the threshold — the reservation and the actual request currently disagree.
opencode-agent commentedon Sep 27, 2026
Fix merged into
v2in #51271. It landed afterv2.0.18and will be included in the next V2 release,v2.0.19. Compaction requests now send an explicit output-token limit fitted to the reserved context room instead of letting the provider choose an oversized default.