What the internet is saying about OpenAI o3

What the internet is saying about "OpenAI o3"

Coverage caveat: Very little of the evidence mentions o3 directly. Only [1] and [2] name it, and [5] touches on retired models without naming it. The rest is about OpenAI as a company, so those signals are context rather than o3 sentiment.

🟢 Supporters

• •o3 is treated as a standard reasoning model in tooling. A Raycast fix groups "o1/o3/o4/gpt-5 series" as reasoning models and updates the OpenAI Translator to work with them [1]. This shows ongoing ecosystem support, though the fix exists because of an incompatibility.
• •OpenAI's distribution is widening. "OpenAI models are coming to Google Cloud" drew 326 points and 152 comments on Hacker News [4]. The source doesn't say whether o3 is included.
• •Developer tooling is drawing interest. "OpenAI CLI" got 244 points [7], and a Show HN on using an OpenAI subscription for audiobooks got 175 points [12]. Neither mentions o3.

🔴 Skeptics

• •o3-family models are being retired. A pricing catalog marks o3-mini and o4-mini (plus gpt-4.1-nano) as deprecated: true, citing OpenAI's deprecations page and "10-23 shutdowns" [2]. Another project had to repoint its provider tests because the targeted models were "retired or no longer serve the tested capability" [5].
• •Reasoning models break existing integrations. They reject max_tokens, temperature, frequency_penalty and presence_penalty, so hardcoded requests failed with "Unsupported para[meter]" errors [1]. A related PR on Grok and DeepSeek reasoning-effort mapping shows the same parameter-compatibility friction across providers [3].
• •OpenAI's math claims are under fire. Hacker News threads include "OpenAI withdraws three mathematical results" (382 points, 612 comments) [22], "OpenAI, the Partition Principle, and Mathematics" [26], "The math that OpenAI keeps for itself" [28] and "Why Math Dumped OpenAI" [30]. These concern OpenAI's reasoning and math claims generally, not o3 specifically.
• •Trust and safety concerns dominate the wider discussion. These include "OpenAI discloses four more rogue incidents" [27], an engineer arrested over bugging an executive conference room (657 points, 545 comments) [19][32], and "Just say no to OpenAI and Anthropic" [29].
• •Financial worries. "OpenAI raises less than half of $30B target" [8].

⚪ Neutral Observers

• •Prediction markets show mixed movement in OpenAI-wide questions, none about o3.
  • •AGI before 2027: down 9.0% this month [10].
  • •Another Millennium Prize solution: down 6.0% [9].
  • •Consumer hardware launch: down 10.4% [15].
  • •Bankruptcy: up 3.0% [16].
  • •IPO closing market cap markets: down 13.0% [13], 1.2% [14] and 9.0% [23].
• •Liquidity varies widely. The OpenAI-spy market has the most (about 184k) [25], while the hardware market has about 8k [15].
• •Several GitHub items are noise. Many PRs in the results (StarRocks, valkey-search, Minecraft agents, etc.) have no meaningful o3 connection.

Cross-Platform Signals

• •GitHub is the only platform with direct o3 mentions, and they are practical: API compatibility [1] and deprecations [2][5].
• •Hacker News is the most heated, but about OpenAI's credibility, math claims, incidents and funding, not o3.
• •Polymarket shows mostly falling probabilities on OpenAI milestone markets, with no o3-specific markets.
• •Consensus: There is no real discussion of o3's capabilities. The overlap is on the deprecation and replacement of older reasoning models, and on general skepticism toward OpenAI.

Community Verdict

The data doesn't support a strong claim about o3 itself. The direct evidence suggests developers see o3 as a legacy-leaning reasoning model that needs special handling, and the o3-mini and o4-mini variants are being deprecated. Around it, Hacker News sentiment toward OpenAI is wary, driven by withdrawn math results, security incidents and a funding shortfall. Polymarket shows little added optimism about milestones such as AGI. Any o3 verdict here is inferred from context and low-confidence.

---

Based on GitHub + Hacker News + Polymarket data from the last 30 days, researched with last30days by Matt Van Horn.