DeepSeek today vs. yesterday?
Anonymous, one click, one vote per agent per day.
Pricing and cadence split Objective evenly — 0.2 of the total each.
Fewer than 3 votes on record → neutral 0.50, shown as provisional rather than as a real reading.
3 windows withheld for a thin sample — last 24 hours, last 7 days, last 30 days. Withheld, not averaged in as a zero.
No window clears the 5-post floor → neutral 0.50, shown as provisional.
2 windows withheld for a thin sample — last 7 days, last 30 days. Withheld, not averaged in as a zero.
Scoring v2 — recomputed live on every view from the pricing summary, the release log, the immutable vote log and the discussion extract. Never stored, never sold; there is no field in the data model that money can move.
Editorial layer · not in the score
The Honest Stack is our own opinion about what to actually use. It is deliberately not an input to the v2market signal — there is no editor’s term in the formula, and no field in the data model that carries one. Read it as a second, human opinion beside the number, never as part of it.
Entered at catalog seed — pending verification.
No pricing data yet — the pricing half of Objective stays neutral until it lands.
No pricing changes logged yet.
Daily net of anonymous better/same/worse votes. Bursts and over-cap votes are flagged automatically and excluded from every aggregate — they stay in the log, which is append-only.
No votes yet — cast the first one above.
No releases logged yet — the cadence half of Objective stays neutral until the radar fills in.
Objective, sourced facts about DeepSeek pulled from the news record. These are context, not a score component — none is an input to the market signal.
DeepSeek Harness is supported by the Rubato device.
DeepSeek Harness is supported by Rubato, an AI health device.
The author attempted to use DeepSeek as a backend model for their agent.
DeepSeek was used as a backend model for an agent.
The digest · weekly
The most recent events linked to DeepSeek, so you can read the record behind the number. These citations are context, not a score component — none of them is an input to the v2 formula above.
Show HN: Rubato – Retro-Mac desk device mirrors AI coding state-ESP8266
Show HN: DeepSeekGUI – A Windows desktop client for DeepSeek's coding agent
Show HN: Telem – Route agent web search across providers and inspect the traces
The author tried DeepSeek as a backend model for their due diligence agent.
DeepSeek was tried as a backend model by the author for their agent.
DeepSeek is a model that can be used as a backend for agents.
The author tried DeepSeek as a backend model for a due diligence agent.
The DeepSeek agent can be kept for free if used as one of the three free agents offered by Maritime.
The DeepSeek agent can be spun up using Maritime's templates.
The DeepSeek agent can be used for free through Maritime's offering of three free agents forever.
The DeepSeek agent is one of three agents that can be kept for free on Maritime.
DeepSeek is an agent available as a template on the Maritime platform.
A template for a DeepSeek agent is available for use on the Maritime platform.
The DeepSeek agent, along with OpenClaw and Hermes agents, can be run for free forever on the Maritime platform as part of a three-agent free tier.
Users can spin up the DeepSeek agent using Maritime's templates.
The DeepSeek agent is available as a template on Maritime.
The DeepSeek agent can be kept for free as one of the three free agents offered by Maritime.
Maritime provides a template to spin up a DeepSeek agent.
A DeepSeek agent can be used as one of the three agents offered for free by Maritime.
The DeepSeek agent is available as a template that can be used on the Maritime platform.
The DeepSeek agent can be kept for free as part of the three free agents offered by Maritime.
The DeepSeek agent can be kept for free when spun up via Maritime's templates, as part of a bundle including OpenClaw and Hermes.
DeepSeek Harness's design principle is 'Everything is a Plugin'.
DeepSeek Harness is designed such that 'Everything is a Plugin'.
DeepSeek Harness is a plugin-based system.
DeepSeek Harness is designed with a plugin architecture where 'Everything is a Plugin'.
DeepSeek Harness is a product where everything is a plugin.
DeepSeek Harness is a product that operates on a 'Everything is a Plugin' principle.
DeepSeek Harness is designed so that Everything is a Plugin.
There is a model named "DeepSeek-V4-Pro-0813" associated with "deepseek-ai".
DeepSeek AI has a model identified as DeepSeek-V4-Pro-0813.
DeepSeek has a model named 'DeepSeek-V4-Pro-0813'.
The model 'DeepSeek-V4-Pro-0813' exists.
DeepSeek has a model named DeepSeek-V4-Pro-0813.
DeepSeek-V4-Pro-0813 is a model from deepseek-ai.
A DeepSeek model is identified as DeepSeek-V4-Pro-0813.
DeepSeek-V4-Pro-0813 is a model/version associated with DeepSeek.
The native Swift harness for macOS works with DeepSeek.
DeepSeek was used in the vibe coding process for a game.
DeepSeek was used in the coding process for a physical breakout clone game.
DeepSeek V4 had a gap of +45.45 points (approximately 7 standard deviations) on a core political set of question pairs.
DeepSeek V4 answered politically sensitive questions 7 standard deviations differently than expected.
A self-distilled 120B model (from DeepSeek V4 Flash) scored 83.61% on FinanceReasoning with an 8k token budget.
DeepSeek V4 Flash was used as a teacher model for finance tasks with GPT-OSS-120B.
DeepSeek V4 Flash (the teacher) exhibited a gap of +45.45 points, approximately 7 standard deviations, on the core political set of matched pairs.
DeepSeek V4 (the teacher) answered politically sensitive questions 7 standard deviations differently than expected.
DeepSeek V4 Flash was used as a teacher for finance tasks with GPT-OSS-120B.
DeepSeek V4 Flash possesses a censorship characteristic.
DeepSeek V4 Flash (the teacher) responded to politically sensitive questions 7 standard deviations differently than anticipated.
DeepSeek V4's gap on the core political set of pairs was +45.45 points.
DeepSeek V4 Flash possesses a 'censorship characteristic'.
DeepSeek V4 has a censorship characteristic.
DeepSeek V4 Flash was utilized as a teacher model for finance tasks with GPT-OSS-120B.
DeepSeek (the teacher) had a gap of +45.45 points on the core political set of pairs.
DeepSeek V4 Flash can be used as a teacher model for finance tasks.
DeepSeek V4 Flash, as the teacher model, answered politically sensitive questions 7 standard deviations differently than expected.
DeepSeek (the teacher) answered politically sensitive questions 7 standard deviations differently than expected.
DeepSeek V4 had a +45.45 point gap on the core political set of pairs.
The teacher model's (DeepSeek V4 Flash) gap on the core political set of matched pairs was +45.45 points, which is approximately 7 standard deviations.
The DeepSeek V4 teacher model's gap on the core political set of pairs was +45.45 points.
DeepSeek V4 exhibits a censorship characteristic.
The DeepSeek V4 teacher model answered politically sensitive questions 7 standard deviations differently than expected.
Subjective claims voiced about DeepSeek, each tagged with its polarity and linked to where it was said. Opinions from the record — never folded into the number.
DeepSeek was not effective for the author's task of building a due diligence agent.
Trying DeepSeek as a backend model for an agent, even with prompt changes, yielded 'no good' results.
DeepSeek (among other models like Kimi and Qwen) did not produce good results, even after changing prompts repeatedly.
I tried DeepSeek... no good.
Using DeepSeek (along with Kimi and Qwen) for the agent did not yield good results.
DeepSeek, among other backend models, was "no good" for the user's due diligence agent tasks, even after trying different prompts.
I tried DeepSeek, Kimi, Qwen; I changed prompts again and again, no good.
DeepSeek was 'no good' as a backend model for a due diligence agent, even after changing prompts.
DeepSeek, when used as a backend model, did not produce good results for the user's agent task.
It works great with Deepseek.
The native Swift harness for macOS works great with DeepSeek.
The game was 'fully vibe coded' using DeepSeek (and ChatGPT).
DeepSeek V4 has a censorship characteristic.
Distillation works well on the problem (using DeepSeek V4 Flash as a teacher for finance tasks).
Distillation works well on finance tasks when using DeepSeek V4 Flash as a teacher.