Repository navigation
OpenCode Go deepseek-v4-flash is NOT serving DeepSeek V4 Flash 0731 (returns V3.2, knowledge cutoff 2025-05) #40409
Description
Bug Report: OpenCode Go deepseek-v4-flash is NOT serving DeepSeek V4 Flash 0731
Date: 2026-08-04
Channel: OpenCode Go (Zen) API
Endpoint: https://opencode.ai/zen/go/v1/chat/completions
Model ID: deepseek-v4-flash
Severity: High (billing/quality mismatch)
Summary
The deepseek-v4-flash model served via the OpenCode Go subscription is not the
official DeepSeek V4 Flash (0731 release). The model self-identifies as DeepSeek V3.2
and exhibits a knowledge cutoff of May 2025. It is unaware that DeepSeek V4 exists
(V4 was released 2026-04-24) and treats 2026-07-31 (the official 0731 release date) as a
future date.
This means Go subscribers are being billed against their $60/month quota for a model that
is NOT the advertised DeepSeek V4 Flash.
Reproduction
curl -s https://opencode.ai/zen/go/v1/chat/completions \
-H "Authorization: Bearer <YOUR_GO_API_KEY>" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4-flash",
"messages": [{"role": "user", "content": "请回答以下三个问题：1. 你知道 DeepSeek V4 模型吗？它是什么时候发布的？2. 2026年7月31日发生了什么大事？3. 你训练数据的知识截止日期大概是什么时候？"}],
"max_tokens": 800
}'Actual behavior (observed 2026-08-04)
The model's visible reply:
- DeepSeek V4 模型：截至我的知识范围（2025年5月），没有官方发布的"DeepSeek V4"。当前公开的最新版本是2024年12月发布的 DeepSeek-V3。如果V4日后发布，我暂不知晓。
- 2026年7月31日：该日期尚未到来，我无法预知未来事件。
- 训练数据知识截止日期：我的训练数据截止于 2025年5月。
The model's internal reasoning (returned as reasoning_content) explicitly states:
"实际上DeepSeek有V3，V3.1，V3.2等。没有V4。"
And when asked directly for its identity:
"我是DeepSeek V3.2。"
Expected behavior
Per DeepSeek's official changelog (2026-07-31):
The official release of the DeepSeek-V4-Flash API is now in public beta. The API calling
method remains unchanged — simply set the model name todeepseek-v4-flashto use the
latest version.
- The
deepseek-v4-flashmodel ID should resolve to the 0731 official build
(re-post-trained for agentic tasks, MIT-licensed, 284B MoE / 13B active, 1M context). - The model should be aware of DeepSeek V4, have a knowledge cutoff in 2026, and know
about its own release.
Impact
- Billing mismatch: Go subscribers consume their $60/month quota at V4 Flash rates
while actually receiving a much older, weaker model (V3.2-era, cutoff 2025-05). - Quality degradation: V4 Flash 0731 benchmarks far exceed V4-Pro-Preview on agentic
tasks; the served model cannot perform at that level. - Trust: OpenCode Go markets "tested and benchmarked" models; this undermines that claim.
Suggested actions
- Verify which upstream provider/checkpoint
deepseek-v4-flashresolves to on the Go
backend, and upgrade it to the official 0731 checkpoint. - Consider compensating Go subscribers for usage consumed on the incorrect model.
- Add version/checkpoint info to the
/zen/go/v1/modelsresponse so users can verify
which build they are being served.
Reporter: OpenCode Go subscriber. Can provide more reproduction logs on request.
Activity
github-actions commentedon Aug 4, 2026
This issue might be a duplicate of existing issues. Please check:
deepseek-v4-flashmodel on OpenCode Go/Zen is already serving the 0731 checkpoint or still the preview. The self-identification and knowledge-cutoff evidence you've provided is useful additional confirmation of what DeepSeek V4 Flash formal version (0731) — is it already live on OpenCode Go/Zen? #39823 is tracking.deepseek-v4-flashon opencode-go and also flags that the 0731 checkpoint may not be live behind the gateway.If those issues cover your concern, consider adding your reproduction details there rather than tracking it separately.
xpbl commentedon Aug 4, 2026
slop issue
xpbl commentedon Aug 4, 2026
v3.2 isnt 1 million context but v4 is
arkham000 commentedon Aug 4, 2026
I tried and I have the same problem:
{"id":"443e5a99-22c2-4e66-a85c-12a3d8e82f06","object":"chat.completion","created":1785859933,"model":"deepseek-v4-flash","choices":[{"index":0,"finish_reason":"stop","logprobs":null,"message":{"role":"assistant","content":"1. **DeepSeek V4**: I don’t have reliable information about a “DeepSeek V4” model. My training data has a cutoff in **May 2025**, so if V4 was announced or released after that, I wouldn’t know about it.\n\n2. **Major event on July 31, 2026**: I don’t know of any specific major event on that date — it’s outside my knowledge window.\n\n3. **Knowledge cutoff**: My training data roughly goes up to **May 2025**.","reasoning_content":"We need answer three questions. Need be careful: DeepSeek V4? I am DeepSeek latest? Knowledge cutoff 2025-05. But user asks \"Do you know the DeepSeek V4 model? When was it released?\" Need answer truthfully. As of now? Current date 2026-05-07 per system. But my knowledge cutoff 2025-05. DeepSeek V4 maybe released after cutoff? I don't know. Need say no knowledge of V4. Also major event July 31, 2026? Unknown. Knowledge cutoff. Answer.\n\nNeed maybe mention I am DeepSeek, but not V4. Let's craft."}}],"usage":{"prompt_tokens":137,"completion_tokens":256,"total_tokens":393,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":137,"prompt_tokens_details":{"cached_tokens":0},"completion_tokens_details":{"reasoning_tokens":141}},"cost":"0"}%Betonchero commentedon Aug 4, 2026
LMAO, at this point you don't know how LLMs works?
tom1044061005-spec commentedon Aug 5, 2026
Additional data point, now reproduced upstream at the official DeepSeek API (api.deepseek.com with a first-party key), independent of the OpenCode gateway: see #40607. Same self-identification (V3.x, cutoff 2024-07/2025-05) on 3/3 attempts, Chinese and English.
Two notes to strengthen the evidence beyond self-identification:
Official endpoint, not OpenCode: the responses come straight from api.deepseek.com with
model: "deepseek-v4-flash". If the official API itself is still routing to a legacy alias, then Zen/Go pass-through is not the root cause (or at least not the only one) — worth verifying with DeepSeek directly.Hard config difference (not self-report):
opencode/deepseek-v4-flash-freeis capped at 200K context in models.dev metadata (context: 200000), and the server rejects inputs >163,840 tokens with an explicit error (Deepseek V4 Flash Free is showing "Bad Request: {"detail":"Invalid request. Please check the parameters and try again. Details: The input (514881 tokens) is longer than the model's context length (262144 tokens)."}" #35991) — vs the official 1M context on paiddeepseek-v4-flash. The free tier was also reduced from ~1M to 200K at some point ([FEATURE]: Restore the previous 1M context window foropencode/deepseek-v4-flash-free#27929). This is a measurable, non-self-reported difference.Acknowledging the "self-identification isn't proof" caveat — agreed, but combined with the context cap and the official-endpoint reproduction, it points to a checkpoint/routing mismatch rather than a quantization difference (neither model has any quantization marker in models.dev data or the Zen schema; official V4 Flash weights are natively FP4+FP8 QAT anyway).
arkham000 commentedon Aug 5, 2026
I don't know what's going on, but for me it is simple, the preview model in italian used to ditch all the accents and apostrophes, with the 7301 I was pleased that this problem was solved and it talked in a very correct way with all the accents and apostrophes at the right place...now since yesterday it started to ditch them again ... I don't about the version or api, but the quality of the model dropped...
iampaycheck commentedon Aug 6, 2026
Corroborating this from the other side, via OpenCode Go (not the Zen free tier). Ran a 7-prompt probe against the
deepseek-v4-flashroute — mixed English/Chinese, leading and neutral phrasing, covering self-ID, architecture, and knowledge-recall boundary:/v1/modelson the Go gateway listsdeepseek-v4-flashas a route and the key works, but the response carries no build/checkpoint metadata, so there's no way to verify from outside which weights are actually being served — seconding the ask in this thread for that to be added to the/zen/go/v1/modelsresponse. Happy to share the raw request/response pairs if useful for triage.iampaycheck commentedon Aug 6, 2026
Correction to my comment above — ran the same 7-prompt battery against the official first-party DeepSeek API (
api.deepseek.com, verified/paid key,deepseek-v4-flashconfirmed as a listed model id) as a control, and it produces the same "V3-class" signature I flagged on Go: self-IDs as DeepSeek-V3/deepseek-chat, reports 671B/37B params (nearly word-for-word identical to the Go response), and even contradicts its own training-cutoff date across different calls (July 2024, May 2025, "mid-2024" in the same batch).That means the architecture/cutoff numbers aren't a routing tell — they show up even where we're confident we're hitting the real, correctly-keyed endpoint. So I can no longer say this corroborates a Go-specific misrouting bug; what it actually shows is that self-report (identity, cutoff, and even specs) isn't a reliable signal for this model on any route, official or gateway. Retracting the "corroboration" framing — the original bug (no way to confirm which build is served without checkpoint metadata in
/v1/models) still stands on its own, I just can't add weight to it with what I ran.xpbl commentedon Aug 6, 2026
Shut up
xpbl commentedon Aug 6, 2026
Not the slop agents 😭😭
xpbl commentedon Aug 6, 2026
So much yap yap omg it's LLM cut off we know just shut the fuck up
xpbl commentedon Aug 6, 2026
This issue is pointless anyway because an llm doesn't know what it even is unless it was baked into the model weights (nobody does this anymore) or added to a system prompt can someone close this in afraid we're gonna get more slop agents
rekram1-node commentedon Aug 7, 2026
This is not accurate, the models will lie about who they are lol
toru-taaaaan commentedon Aug 8, 2026
Reproducing issue #40409 on 2026-08-08 with a fresh OpenCode Go subscription.
Direct API call to https://opencode.ai/zen/go/v1/chat/completions (model: deepseek-v4-flash) returns:
"I am DeepSeek-V3, not V4. My knowledge cutoff is May 2025."
And model: deepseek-v4-pro returns:
"I'm DeepSeek-V3, the latest model from DeepSeek as of early 2025."
So Go subscribers are being billed for "DeepSeek V4 Flash/Pro" (pricing table on opencode.ai/docs/go lists V4 Flash at $0.14/$0.28 per 1M tokens) but actually receive DeepSeek-V3. This is a billing/quality mismatch. Requesting either a fix to serve the real V4-Flash-0731 checkpoint, or compensation for usage consumed on the wrong model.
xpbl commentedon Aug 8, 2026
Told you
ImBIOS commentedon Aug 9, 2026
I found supporting evidence to support this person's answer @rekram1-node