Denser vs Wonderchat
Wonderchat cites its replies and flags weak answers, and on those it matches Denser. It does not publish how retrieval works, and it holds 1,000 webpages per agent against Denser's 2,000 pages on the equivalent tier and 50,000 on Enterprise.
2,000+ businesses · free plan, no card required · 5-minute setup
Trusted by 2,000+ businesses worldwide
City of San José
Onondaga County
EOS Australia
Portfoliobox
Upsiide
Valk WeldingCity of San JoséOnondaga CountyEOS AustraliaPortfolioboxUpsiideValk WeldingA chatbot can only answer from the passage retrieval hands it, so retrieval quality sets the ceiling on everything downstream. Denser Retriever combines keyword search, vector search, and an XGBoost reranker, and on MTEB retrieval benchmarks that combination beats a vector-search baseline outright, by 13.07% NDCG@10 over top vector baselines on MS MARCO. Wonderchat describes a grounded retrieval engine but never publishes how it works, so its accuracy cannot be independently checked. In support, that gap is the difference between a correct answer and a confident wrong one.
A Denser bot holds 2,000 pages or 1 GB on Standard and 50,000 pages or 25 GB on Enterprise. Wonderchat holds 1,000 webpages per agent on Basic and 2,500 on Turbo, which is a help centre rather than a documentation estate.
Both products cite their sources, so the question is what sits underneath the citation. Denser Retriever is MIT-licensed on GitHub and you can read exactly how a passage is selected; with Wonderchat you are trusting a description of the mechanism.
Denser grades every answer and clusters the failures by topic, so gaps in your content arrive as a ranked list rather than as complaints.
Wonderchat has native AI-to-human handover that passes full conversation context to an agent, where Denser handles escalation through actions and integrations, and it offers an on-premise deployment option. Both matter if escalation or self-hosting is central to your workflow. Neither changes retrieval quality, content capacity, or cost per answer, and on that last one Denser answers roughly seven times as many questions per dollar at the comparable tier.
| Denser | Wonderchat | |
|---|---|---|
| Retrieval | Hybrid keyword + vector + XGBoost reranking, MTEB-benchmarked above vector search alone. Open source (MIT) | Grounded retrieval engine, not disclosed |
| Content per bot | 100 pages (Starter) to 50,000 / 25 GB (Enterprise) | 1,000 webpages (Basic), 2,500 (Turbo) |
| Entry paid plan | $39/mo, 1,500 queries | $124/mo (annual), ~1,000 resolutions |
| Mid tier | $119/mo, 7,500 queries | $417/mo (annual), ~5,000 resolutions |
| Free plan | 20 queries/mo, ongoing | ~10 resolutions/mo, 50 webpages |
| Billing unit | 1 query per question | Credits, ~5 per resolution |
| Enterprise entry | $899/mo published | From $1,499 |
| Source citations | Every answer cited to its exact passage | Yes — every reply cited |
| Weak-answer flagging | Flags unanswered questions by cause and topic | QA agent flags weak answers |
| Model choice | Latest Claude, GPT, and Gemini, all at one query each | OpenAI, Anthropic, Google, Mistral |
| Deployment | Widget, iframe, React, REST API, TS/Python SDKs, WordPress, Shopify, Slack, WhatsApp | Widget, integrations, API |
| Human handover | Actions and integrations | Native, with full context |
| On-premise option | Not offered | Available |
| Best for | Cost-efficient answers over a large knowledge base | Support flows that escalate to humans |
Real cost
Wonderchat meters resolutions and Denser meters questions, so this assumes a five-question conversation — roughly 1,000 Wonderchat resolutions against 5,000 Denser queries.
| Denser | Wonderchat | |
|---|---|---|
| Plan needed | Standard, $119/mo (7,500 queries) | Basic, $124/mo annual (~1,000 resolutions) |
| Headroom | 2,500 queries | None at the assumed conversation length |
| If conversations run longer | No change until 7,500 questions | Credits burn faster; Turbo is $417/mo |
| Extra volume | Move to the next tier | $29.99 per extra 1,000 credits |
| Monthly total | $119 | $124, or $417 at Turbo |
Rates read from wonderchat.io/pricing on 19 August 2026, on annual billing. Wonderchat meters resolutions and Denser meters questions, converted here at five questions per conversation.
Switching
Wonderchat trains on your help centre, docs, and past tickets. Denser takes the same sources, so the migration is a re-index rather than an export.
Both products cite. Ask the same question of each and check which one points at the better passage — this is the rare comparison where you can judge retrieval quality side by side.
If you relied on Wonderchat's native handover, decide how escalation works in Denser through actions and integrations before you cut over. This is the real switching cost.
Pull last month's conversation count and average length, then price both. Once resolutions are converted into questions, the gap in cost per answer is straightforward to see.
Questions
Yes, particularly on cost and content scale. Both cite every answer and flag weak ones, so the decision usually comes down to throughput per dollar — Denser Standard is $119/mo for 7,500 queries against Wonderchat Basic at $124/mo for roughly 1,000 resolutions — and to how much content one bot must hold. Wonderchat stays ahead on native human handover and its on-premise option.
Yes, prominently. Wonderchat states that every response cites its source and that its engine retrieves, grounds, and cites every reply. Citations are not the differentiator between these two, so the decision comes down to cost per answer, content per bot, and whether the retrieval stack is open to inspection.
Denser, at any meaningful volume. Wonderchat's first paid tier is $124/mo on annual billing and its enterprise plan starts at $1,499; Denser's first paid tier is $39/mo and its top published plan is $899/mo. The units differ — Wonderchat meters resolutions, Denser meters questions — so price your own transcript volume before deciding.
A Wonderchat resolution covers a whole conversation and consumes roughly five credits. A Denser query is one customer question. A conversation of five questions is therefore about one Wonderchat resolution and five Denser queries, which is the conversion to apply when comparing the two price lists.
Through actions and integrations rather than a native agent inbox. Wonderchat's handover is more complete: it passes full conversation context to a human agent as a first-class feature. If escalation is central to your support process, that gap is the strongest argument for Wonderchat.
Yes. Denser supports the latest Claude, GPT, and Gemini models, and every one of them costs the same single query per customer question. Wonderchat also offers model choice and adds Mistral alongside an on-premise option, so the difference is not whether you can pick a stronger model but whether picking one changes your bill. On Denser it does not.
Point Denser at your website or upload your documents, ask the questions your customers actually ask, and check the source behind every answer. It takes about five minutes.
