Enterprise Infrastructure — Hosted on argonnetwork.biz.id
Enterprise Model Catalog
Access reasoning, MoE, and agentic foundation models from Z.ai, Reflection AI, Anthropic, OpenAI/GPT, DeepSeek, and Agnes AI through a single SOC 2 compliant gateway on argonnetwork.biz.id. Switch models with a string, governed by one master agreement.
Catalog specifications
All active models deployable through https://api.argonnetwork.biz.id/v1. Exact pricing, latency benchmarks, context windows, and evaluation availability.
| Model ID | Provider | Family / Arch | Context | Input / 1M | Output / 1M | Latency | Operational Status | Evaluation Tier |
|---|---|---|---|---|---|---|---|---|
|
glm-5.3-flash GLM 5.3 Flash |
Z.ai | Reasoning & Thinking | 500K | $0.00 | $0.00 | 0 ms | Available | Unlimited Trial |
|
beam-1.0 Beam 1.0 |
Reflection AI | Sparse MoE (501B) | 1M | $0.00 | $0.00 | 380 ms | Maintenance | Unlimited Trial |
|
claude-sonnet-4-6 Claude Sonnet 4.6 |
Anthropic | Frontier Intelligence | 270K | $3.00 | $15.00 | 380 ms | Available | Unlimited Trial |
|
gpt-6-luna GPT 6 Luna |
GPT | Codex & Deep Reasoning | 270K | $2.00 | $2.00 | 380 ms | Available | Unlimited Trial |
|
deepseek-v4.1 DeepSeek V4.1 Flash |
DeepSeek | High Throughput MoE | 260K | $0.15 | $0.60 | 380 ms | Available | Unlimited Trial |
|
agnes-3.0-flash Agnes 3.0 Flash |
Agnes AI | Agentic Orchestration | 512K | $0.05 | $0.15 | 186 ms | Available | Unlimited Trial |
|
auto Auto (policy routing) |
Argon Digital | Dynamic Routing Layer | Dynamic | Variable | Variable | 35 ms | Available | Unlimited Trial |
|
deepseek-4.1-flash-1017 DeepSeek 4.1 Flash 1017 |
DeepSeek | Snapshot Long Context | 1M | $0.30 | $1.20 | 210 ms | Unlock | Lite / Pro |
|
gpt-5.6-sol GPT 5.6 Sol |
GPT | Priority Access | 270K | $4.00 | $20.00 | 320 ms | Unlock | Lite / Pro |
|
gpt-astra GPT 6 Astra |
GPT | Frontier Flagship | 270K | $10.00 | $50.00 | 420 ms | Maintenance | Lite / Pro |
|
gpt-5.6-terra GPT 5.6 Terra |
GPT | Professional Code | 256K | $2.50 | $15.00 | 340 ms | Unlock | Lite / Pro |
|
claude-sonnet-4-6-pro Claude Sonnet 4.6 Pro ∞ |
Anthropic | Priority Pro | 270K | $3.00 | $15.00 | 280 ms | Unlock | Lite / Pro |
Context allowances represent combined input and generation tokens where supported upstream. Output latency values reflect median TTFT (Time To First Token) recorded across global Argon Digital edge clusters over the trailing 24 hours. The auto router dynamically selects the optimal compliant model; actual billing matches the executed upstream model at parity.
Available in Unlimited Trial
Institutional evaluation workspaces receive instantaneous, zero-cost access to these 6 flagship models plus our intelligent Auto routing policy. Fully governed, unmetered request limits, no credit card required.
glm-5.3-flash
GLM 5.3 Flash flagship open-source reasoning model. 500K context, thinking & tool calling.
beam-1.0
Reflection AI's flagship 501B sparse MoE model (23B active). 1M context, agentic tool use, SWE-bench coding & STEM reasoning.
claude-sonnet-4-6
Strong balance of speed and intelligence. Reliable for most enterprise workloads.
gpt-6-luna
OpenAI's latest GPT-6 Luna flagship model running in Codex. 270K context, high-speed generation, deep reasoning, and full tool use.
deepseek-v4.1
Next-generation DeepSeek flagship model with advanced MoE reasoning, 270K context, and ultra-fast high-throughput inference. Unlimited tokens, free for Lite/Pro.
agnes-3.0-flash
Agnes AI next-gen flagship agentic model. 512K context, fast tool orchestration and reliable delivery. Unlimited tokens, free for Lite/Pro.
auto
Routes each request across your compliant active models based on latency targets, cost ceilings, and residency constraints.
Lite / Pro Tier Models
Specialized high-capacity models available for instant commercial unlock or production commitments. Designed for high-concurrency coding, long-context document analysis, and dedicated token allocations.
deepseek-4.1-flash-1017
DeepSeek 4.1 Flash snapshot (1017) with 1M context, ultra-fast generation, coding & reasoning. Available for Lite/Pro members.
gpt-5.6-sol
Premium GPT 5.6 model for Lite/Pro users. 270K context, enhanced performance, priority access, and full tool use support.
gpt-astra
OpenAI's frontier flagship model running in Codex. 270K context, state-of-the-art reasoning, advanced agentic coding, and full tool use support. Free for Lite/Pro: 25M tokens/day, no balance used.
gpt-5.6-terra
Balanced GPT model. Optimized for professional workflows and code.
claude-sonnet-4-6-pro
Premium Claude Sonnet 4.6 for Lite/Pro users. Priority access and faster responses.
Enterprise Routing Layer
Targeting auto relinquishes manual model selection to Argon Digital's deterministic arbitration engine. Every incoming request is scored against latency constraints, residency jurisdictions, error budgets, and cost ceilings.
If an upstream provider experiences elevated latency or downtime (such as maintenance intervals), the gateway immediately shifts requests to the next compliant foundation model within milliseconds.
x-argon-model headerDynamic Arbitration Policy
Enterprise Lifecycle Management
We provide a minimum of ninety (90) days' formal written advance notice before any model identifier or alias is phased out from the institutional gateway.
Explicitly pinned model versions continue to serve existing routing policies through published deprecation dates, protecting enterprise workflows from unannounced shifts.
Trial workspaces execute on the exact same production inference clusters, SLAs, and security perimeters as commercial enterprise deployments.
Institutional Access
Initiate your 30-day unlimited evaluation across our 6 trial foundation models. Provisioned on a dedicated gateway at api.argonnetwork.biz.id.