Xiaomi MiMo-V2-Pro
xiaomi/mimo-v2-pro
1M context
Xiaomi positions MiMo-V2-Pro as a global top-tier agent model. In practice, it gives free users a genuinely large-context partner option for long-horizon coding, planning, and agent loops.
SourceFrontier Model Refresh
We just moved a chunk of the frontier into the free tier.
This launch is about more than adding model IDs. We re-cut the lineup around what users actually do: long-context coding, screenshot-heavy debugging, multimodal planning, and agent loops that keep running after the first response.
Free frontier
3
MiMo V2 Pro, GLM-5V Turbo, and Gemma 4 31B IT now ship as free partner models.
Premium refresh
3
GPT-5.4, GLM-5, and MiniMax M2.7 now anchor the paid frontier tier.
Biggest jump
1M
Both GPT-5.4 and MiMo V2 Pro push the active catalog into million-token territory.
Free partner models
Z.AI GLM-5V-Turbo
z-ai/glm-5v-turbo
202K context
GLM-5V-Turbo is the free multimodal agent pick. It is built for image, video, and text inputs, which matters when workflows start from screenshots, UI bugs, diagrams, and documents instead of plain prompts.
SourceGoogle Gemma 4 31B IT
google/gemma-4-31b-it
262K context
Gemma 4 31B IT brings dense multimodal reasoning, native function calling, and strong document understanding into the free tier. It is the pragmatic open-weight choice for structured work.
SourcePremium frontier models
OpenAI GPT-5.4
openai/gpt-5.4
1M context
GPT-5.4 is now the flagship premium OpenAI slot in LeemerChat. OpenAI explicitly recommends it for complex reasoning and coding, and the 1M-token window changes what counts as a single-session problem.
SourceZ.AI GLM-5
z-ai/glm-5
80K context
GLM-5 replaces the older GLM-4.7 line across our active configs. The point is not just a newer name: it is the stronger frontier alternative for coding, agent planning, and synthesis-heavy work.
SourceMiniMax M2.7
minimax/minimax-m2.7
200K context
MiniMax positions M2.7 as its next-generation productivity model. For us, it becomes the single MiniMax recommendation for coding, productivity, and tool-oriented execution instead of a split M2.1 versus M2.5 story.
SourceWhy These Count As Frontier
Frontier is now about operating range, not just benchmark bragging.
Frontier no longer means premium-only
The biggest change is economic, not cosmetic. Free users now get serious frontier-grade options instead of fallback models that only make sense for lightweight chat.
The lineup is simpler
We removed overlapping GPT-5 chat variants, retired the GLM-4.7 split, and collapsed MiniMax into one active premium recommendation. Fewer tiers, less selector noise, clearer picks.
Agentic work is the organizing principle
These additions are here because they are good at execution loops: repo reading, long planning arcs, multimodal debugging, and document-grounded reasoning.
Launch Offer
Claim your first month for $1
We start from a base price of $1 and localize the display with a live USD conversion. If you are in Ireland, you will see euro. If you are in India, you will see rupees.
Localized for your regionBase USD fallback
Start with $1
Exchange rate data via ExchangeRate-API. Offer display adapts at runtime based on detected location.
Source notes
This post is based on provider docs, model cards, and public model pages. Where a provider’s public page makes a high-level positioning claim rather than exposing a full benchmark sheet, we phrase that cautiously in the copy above.