Payment ops
For payment reconciliation, our top pick is DeepSeek V4 Pro from DeepSeek. Strong reasoning, a 1M-token context for whole settlement files, and an MIT license you can self-host without legal review. GLM-5.3 is a close second, and Qwen3.8 27B rounds out the list.
Reconciliation agents match processor settlements, bank statements, and your ledger, then explain every break. The work is long, numeric, and touches sensitive money data, so careful reasoning, a long context window, and a license you can self-host under matter most.
Updated
Strong reasoning, a 1M-token context for whole settlement files, and an MIT license you can self-host without legal review.
The official DeepSeek V4 Pro release, with stronger agentic skills and an MIT license.
Input: text
Harnesses in Runtime
Top open-weight agentic scores, so it reliably chains ledger queries, processor APIs, and file parsing across a long run.
Z.ai's strongest open-weight model, post-trained for complex coding and long-horizon agent work.
Input: text
Harnesses in Runtime
Small enough to run on your own GPUs under Apache 2.0 when ledger data cannot leave your VPC.
A compact dense vision-language model under Apache 2.0.
Input: text, image, video
Harnesses in Runtime
Turn the steps your team already follows into a skill the agent reads before every run.
Run two or three of these models against past cases. Evals show which one passes at the lowest cost.
The agent works on its own computer in your VPC, with role-based access and an approval before sensitive actions.
For payment reconciliation, our top pick is DeepSeek V4 Pro from DeepSeek. Strong reasoning, a 1M-token context for whole settlement files, and an MIT license you can self-host without legal review. GLM-5.3 is a close second, and Qwen3.8 27B rounds out the list.
Careful multi-step reasoning over numbers. Long context for full settlement files. Permissive license so ledger data can stay in your cloud. Test two or three candidates against your own past cases before you commit, and keep a human approval on any step that moves money or changes a customer's status.
Yes. Download the weights and serve them with vLLM or SGLang in your own cloud, or use Amazon Bedrock or Google Vertex AI in your account. With Runtime, the agent's computer also runs in your VPC, so card data and PII stay in your environment.
DeepSeek V4 Pro (MIT), Qwen3.8 27B (Apache 2.0) are permissive and fine for commercial use. GLM-5.3 (GLM-5.3 License) uses a custom license, so read the terms first.
