Baidu · Available now
ERNIE-Light-DT-1.0 price per reply in 2026: what it costs and what it does
ERNIE-Light-DT-1.0 is Baidu’s low-cost long-context text model on its Qianfan platform, listed on OpenRouter in November 2026 with a 1,024,000-token context window, at $1.13 per million input and $3.94 per million output tokens there.
Use it now
Chat
≈ 2 credits per reply · See all models
New accounts get 5 free credits once, valid for 7 days. They work on the lower-priced models here; the priciest ones need credits you bought.
Start chatting — 5 free credits- Maker
- Baidu
- Type
- Text
- Released
- No date given
- OpenRouter id
- baidu/ernie-light-dt-1.0-intl
Per million tokens
Price
| - | Input | Output | Source |
|---|---|---|---|
| Via OpenRouter (baidu/ernie-light-dt-1.0-intl) | $1.13 | $3.94 | openrouter.ai |
Checked
Facts, with sources
- Baidu’s Qianfan model page for ERNIE-Light-DT-1.0 does not give a release date; it says the model launched at the same time as ERNIE-5.5-Flash-Preview. We found no dated Baidu announcement.Source: cloud.baidu.com · checked 2026-11-17
- OpenRouter’s endpoint data for baidu/ernie-light-dt-1.0-intl shows a creation timestamp of November 10, 2026 (17:21 UTC) and an endpoint named baidu/ernie-light-dt-1.0-intl-20261110. This is the listing time on OpenRouter, not a Baidu release date.Source: openrouter.ai · checked 2026-11-17 · second-hand
- Baidu describes ERNIE-Light-DT-1.0 as a low-cost long-context text model for long-document Q&A, summarization, multi-turn chat, agent tool calls and batch text processing. Baidu says it ranks in the first tier on public benchmarks for math, reasoning, agents and coding (the maker’s own claim, no figures given, not tested by us).Source: cloud.baidu.com · checked 2026-11-17
- Baidu’s model page: text in, text out; built on Baidu’s "lightning" model work with a sparse attention mechanism called BSA (Baidu Sparse Attention); maximum output 128K; thinking mode can be switched on or off. The page gives no numeric context window.Source: cloud.baidu.com · checked 2026-11-17
- OpenRouter lists a 1,024,000-token context window and up to 131,072 completion tokens, with text input and text output. Supported parameters include reasoning, reasoning_effort, tools, tool_choice, response_format and structured_outputs; implicit caching is supported. Baidu’s pages that we read do not state the 1,024,000 figure (Baidu’s "128K" output matches 131,072 if K means 1,024).Source: openrouter.ai · checked 2026-11-17 · second-hand
- Baidu’s Qianfan price list (in RMB, pay-as-you-go, no length or time tiers): ERNIE-Light-DT-1.0 costs 0.008 yuan per 1,000 input tokens (8 yuan per million), 0.0008 yuan per 1,000 cache-hit input tokens, and 0.028 yuan per 1,000 output tokens (28 yuan per million). Batch inference is not offered. The page lists no USD price and does not say whether it applies to the international listing.Source: cloud.baidu.com · checked 2026-11-17
- OpenRouter lists $1.13 per million input tokens, $3.94 per million output tokens and $0.113 per million cached input tokens (checked November 17, 2026). Baidu is the only host listed.Source: openrouter.ai · checked 2026-11-17 · second-hand
- On OpenRouter the model id is baidu/ernie-light-dt-1.0-intl and its only host is Baidu (Qianfan). The id ends in -intl, which we read as the international listing; neither OpenRouter’s pages nor Baidu’s pages that we read define the suffix. The OpenRouter page says EU in-region routing is not available for this model.Source: openrouter.ai · checked 2026-11-17 · second-hand
- Baidu’s international Qianfan page describes OpenAI-compatible APIs and links to an international console; it does not mention ERNIE-Light-DT or list models, so we could not confirm from Baidu’s international page that this model is offered there.Source: intl.cloud.baidu.com · checked 2026-11-17
- OpenRouter’s description says the model is based on open-source GLM models and includes research from Baidu Search’s Lightning model; Baidu’s model page does not mention GLM. Neither page we read states open weights or a licence, and we did not find downloadable weights.Source: openrouter.ai · checked 2026-11-17 · second-hand
Nearest available
Models you can use now instead
GLM 5.3
Z.ai$0.08 / $3.6 (per million tokens)
- A secondary report (DataNorth) says Z.ai released GLM-5.3 on August 14, 2026; Z.ai’s docs page calls it "Z.ai’s latest flagship model" but states no date.Source: datanorth.ai · checked 2026-11-15 · second-hand
- Z.ai’s docs say GLM-5.3 advances complex software engineering and agent capabilities, with an emphasis on coding and cybersecurity performance.Source: docs.z.ai · checked 2026-11-15
- Context window 1M tokens, maximum output 128K tokens; text-only input.Source: docs.z.ai · checked 2026-11-15
Tencent Hy4
Tencent$0.42 / $1.2505 (per million tokens)
- Tencent’s Hy4 announcement, dated November 4, 2026, calls Hy4 its "latest-generation large language model". Compared with Hy4 Preview it adds visual understanding (a first for the Hy line) and computer use, and Tencent says it improves instruction following, long-context comprehension and long-horizon agent tasks. The maker’s own figures, not tested by us: hallucination rate in Tencent’s internal tests fell from 12.95% to 7.2%.Source: tencent.com · checked 2026-11-17
- Tencent published the Hy4 launch announcement on November 4, 2026, as the full release that follows Hy4 Preview. The announcement does not state a parameter count or a context length for Hy4.Source: tencent.com · checked 2026-11-17
- Hy4 Preview, the earlier version, was announced by Tencent on August 28, 2026 as open source, with 770B total and 49B active parameters and a context window "exceeding 1M tokens". Tencent’s GitHub page for it shows an Apache 2.0 licence badge. These are the preview’s figures; we found no Tencent page that gives the same numbers for Hy4.Source: github.com · checked 2026-11-17
FAQ
Questions
What is ERNIE-Light-DT-1.0?
ERNIE-Light-DT-1.0 is a text model from Baidu. Baidu’s Qianfan page describes it as a low-cost long-context model for long-document Q&A, summarization, multi-turn chat, agent tool calls and batch text processing, with a thinking mode that can be switched on or off. Baidu does not give a release date on that page; OpenRouter listed it on November 10, 2026 (UTC). You can chat with it on this page on credits, without a Baidu account.
How much does ERNIE-Light-DT-1.0 cost?
Baidu’s Qianfan price list is in RMB: 0.008 yuan per 1,000 input tokens and 0.028 yuan per 1,000 output tokens (8 and 28 yuan per million), with cache hits at 0.0008 yuan per 1,000 tokens. OpenRouter lists $1.13 per million input tokens and $3.94 per million output tokens (checked November 17, 2026). Baidu’s page gives no USD price, so we do not convert. On this site a typical reply to this model costs about 2 credits (a credit is about $0.01 in the $5 pack), billed by its real token cost, at least 1 credit.
How do I use ERNIE-Light-DT-1.0?
Through Baidu’s Qianfan platform (API), on OpenRouter as baidu/ernie-light-dt-1.0-intl (Baidu is the only host listed), or in the chat box on this page. We could not confirm from Baidu’s international Qianfan page that the model is offered there.
What is the ERNIE-Light-DT-1.0 context window?
OpenRouter lists a 1,024,000-token context window and up to 131,072 output tokens, text in and text out. Baidu’s model page does not state a context number, and gives a maximum output of 128K. We have not tested the limits.
What does -intl mean in baidu/ernie-light-dt-1.0-intl?
It is the OpenRouter id of the model, and we read the suffix as the international listing. Neither OpenRouter’s pages nor Baidu’s pages that we read explain it, and Baidu’s own model name is ERNIE-Light-DT-1.0. Baidu’s price page is in RMB and does not say whether it applies to this listing.
Is ERNIE-Light-DT-1.0 open source?
We could not confirm it. Neither Baidu’s model page nor OpenRouter’s data states open weights or a licence, and we did not find downloadable weights. OpenRouter’s description says the model is based on open-source GLM models; Baidu’s page does not mention GLM.
Does ERNIE-Light-DT-1.0 have a thinking mode and tool calling?
Baidu’s page says the thinking mode can be switched on or off. OpenRouter lists support for reasoning and reasoning_effort, tools and tool_choice, and JSON-schema structured outputs. We have not tested these.
Is this the official Baidu website?
No. This is an unofficial page, not affiliated with or endorsed by Baidu. Facts link to Baidu’s own pages where possible.