Home / Poe
Quora's multi-model hub — GPT, Claude, Gemini and thousands of community bots behind one points-based subscription
The consolidation play — every major model in one chat for $19.99 — the catch is a points meter whose burn rate depends on the models you choose.
Poe is one points-based subscription for GPT, Claude, Gemini and thousands of bots — consolidation and comparison, not agent execution.
You would otherwise pay for two or more native assistant subscriptions, or you want to test top models side by side before choosing one.
You need agent workflows, deep integration with your documents and apps, or a fixed message allowance you can budget exactly.
Workflows are built only from verified capabilities in our listing research — no vendor copy, nothing invented.
Starts A proposal draft or product description matters enough to compare
You get A best-of-three draft and a model preference you can act on
Model choice becomes evidence, not brand loyalty
Starts A repeat job — SEO titles, spreadsheet formulas, a coding pattern — has a community bot built for it
You get Specialist output without building anything
Other people's prompt engineering becomes your shortcut
Starts Clients or your audience ask the same questions about your offer
You get A shareable bot that answers in your framing
Your most repeated explanation stops costing your time
Starts A post needs both words and a visual
You get Copy and visual from one subscription
No separate image-tool subscription for occasional visuals
Run your real weekly jobs across two or three model families and note which you would actually miss.
The $19.99 tier with 660,000 points/month is the headline plan; the $4.99 Starter suits only light use.
Point costs vary by model — a month of real use tells you whether Premium's pool fits before you consider a higher rung.
Time to first useful result: Under 1 hour · no-code 5/5
Free tier; paid plans are a points ladder — $4.99/month entry, $19.99/month headline plan (660,000 points/month) as the realistic solo plan, up to $249.99/month; yearly billing saves ~17%.
$4.99/month; $49.99/year ($4.17/month billed yearly)
Per credit bundle · $4.99/mo for one person
$19.99/month; $199.99/year ($16.67/month billed yearly)
Per credit bundle · $19.99/mo for one person
$49.99/month; $499.99/year ($41.67/month billed yearly)
Per credit bundle · $49.99/mo for one person
$99.99/month; $999.99/year ($83.33/month billed yearly)
Per credit bundle · $99.99/mo for one person
$249.99/month; $2,499.99/year ($208.33/month billed yearly)
Per credit bundle · $249.99/mo for one person
| Plan | Published price | Solo monthly | Billed by | Limits & overage |
|---|---|---|---|---|
| Starter | $4.99/month; $49.99/year ($4.17/month billed yearly) | $4.99/mo | Per credit bundle | Allowance is a daily points pool, not monthly |
| PremiumSolo pick | $19.99/month; $199.99/year ($16.67/month billed yearly) | $19.99/mo | Per credit bundle | Frontier and reasoning models burn points far faster than standard chat |
| Premium Plus | $49.99/month; $499.99/year ($41.67/month billed yearly) | $49.99/mo | Per credit bundle | Same per-model point costs — the pool is bigger, not cheaper per message |
| Pro | $99.99/month; $999.99/year ($83.33/month billed yearly) | $99.99/mo | Per credit bundle | At this price, compare against subscribing to two native plans directly |
| Pro Max | $249.99/month; $2,499.99/year ($208.33/month billed yearly) | $249.99/mo | Per credit bundle | Priced for maximum throughput — beyond most solo needs |
The ladder itself is smooth ($4.99 to $249.99 in five steps); the risk is inside the meter — point costs vary by model, so switching your daily work to a frontier reasoning model can drain the same monthly pool several times faster without any price change.
Price record · 2026-11-06: Checked 2026-11-06 on poe.com/subscription_plans: the page was read directly this run in its billed-yearly view (SAVE 17%) — $49.99/year for 10,000 points/day; $199.99/year for 660,000 points/month; $499.99/year for 1.65 million points/month; $999.99/year for 3.3 million points/month; $2,499.99/year for 8.25 million points/month. The billed-monthly figures ($4.99, $19.99, $49.99, $99.99, $249.99) are the same page's monthly toggle and are corroborated by multiple dated October-November 2026 reads of the same official page. The free tier's point allowance is not stated on Poe's own subscription page and is recorded as unstated, not estimated.
The same fields on every deep listing in this job, so agents compare field by field. “Not verified” means our research has not confirmed it yet — it never means “doesn't have it.”
Poe solves a specific solo-founder problem: model FOMO. Instead of paying separately for a ChatGPT subscription, a Claude subscription and a Gemini subscription — and still missing the image, video and specialist models — you pay Poe once and reach models from OpenAI, Anthropic, Google and others, plus thousands of community-built bots, through one chat interface and one bill. The metering is the honest trade. Every message spends compute points from your plan's allowance, and the point cost depends on the model you pick, so a month of light chat on efficient models and a month of frontier-model marathons drain the same allowance at very different speeds. Poe shows the point cost as you work, which makes the system predictable in practice even though no plan converts cleanly into a message count. The ladder is unusually granular for this category: a $4.99 entry tier with a daily points allowance, a $19.99 headline tier, and steps up to $249.99 for very heavy multi-model use, each about 17% cheaper billed yearly. For a founder, the fit is comparison and consolidation — test which model writes your proposals best, generate an image in the same session, then settle into the two or three models you actually use, all without three renewals. What Poe is not is an agent platform: bots chat, generate and answer, but Poe does not run your inbox, schedule your week or execute multi-step business workflows the way the agent tools elsewhere in this job do.
Solo Fit stays the public score: the qualitative roll-up of these four published inputs under the How We Rank method. Multi-model consolidation carries it; pricing clarity reflects a published points pool whose per-message cost varies by model.
The difference: The native OpenAI experience with custom GPTs, scheduled tasks and Codex — the catch is it is one provider's models only.
The difference: The native Anthropic experience with Projects and careful long-document writing — the catch is, again, a single provider's models.
The difference: A marketing-content specialist with brand voice and campaigns rather than a model hub — the catch is a $59+ per-seat price for a narrower job.
Paid plans run $4.99/month (10,000 points/day), $19.99/month (660,000 points/month — the realistic solo plan), $49.99, $99.99 and $249.99/month, each about 17% cheaper billed yearly. The directory's True Solo Cost is $4.99 low and $19.99 typical.
There is no fixed conversion. Every model charges its own point cost per message, shown as you work, so efficient models stretch the pool and frontier reasoning models drain it quickly. That model-dependence is exactly why this listing records points, not an estimated message count.
The subscription page read on 2026-11-06 displays its billed-yearly view ($49.99 to $2,499.99 per year, saving 17%). The monthly figures are that same page's billed-monthly toggle, corroborated by multiple dated October-November 2026 reads of the official page — the corroborated-dated-reads treatment this directory uses when a view does not extract cleanly.
For chat and generation across models, it can be. For native ecosystems — ChatGPT's custom GPTs and scheduled tasks, Claude's Projects, Gemini inside your Google apps — those features live in the native subscriptions, not in Poe.
Poe is not in a published stack yet — stacks are edited buying guides, and memberships are added only when the combination has been worked through.
How we rank → · Get a weekly shortlist →