vs
Subconscious vs xAI
Grok doubles the price of any request past 200K tokens. That is exactly where Subconscious is built to run, billing processed tokens with a 5M+ effective context.
By The Subconscious Team · Updated
Subconscious vs xAI: key differences
The pricing rule decides this one. Grok 4.6 costs $2 in and $6 out per million under 200K prompt tokens, but once a prompt crosses 200K, the whole request bills at double. For a long-running agent, 200K is not an edge case. It is where most of the run happens. Grok 4.6 also tops out at 500K, with the older 4.20 and 4.3 models reaching 1M at $1.25 in and $2.50 out. Subconscious has no long-context penalty to dodge. It bills tokens processed after KV cache pruning, so a 1M-token request might bill for 200K, and it delivers a 5M+ effective context window with 2x faster task completion.
Grok has a hook Subconscious cannot match: native Web Search and X Search tools that pull live posts from X, which suits news, market and social-sentiment agents. xAI also sells separate image, video and audio APIs, and its output tokens undercut OpenAI and Anthropic at comparable tiers, so short reasoning and tool-calling work is cheap. The trade is closed Grok weights on a first-party API with a smaller enterprise footprint. Subconscious serves open models through a managed API or dedicated and on-prem deployments, and records no prompts or inputs.
What Subconscious and xAI do
Subconscious
Subconscious is an MIT CSAIL spinout in Kendall Square that builds inference for long-horizon agents, the workloads where a single trace runs past 200K tokens and often into the millions. Its runtime drops in as a replacement for vLLM or SGLang. Instead of rereading an ever-growing context on every step, it prunes the KV cache and preserves suffix state, and Subconscious co-designs the runtime with post-trained model variants it calls Marathon. Against open models on standard inference, Subconscious delivers 2x faster task completion, delivers a 5M+ effective context window, cuts cost 50% and up to 80%, and scores neutral to 10% better on agentic benchmarks.
Example models: GLM 5.3, DeepSeek V4.1 Flash
Full Subconscious profilexAI
xAI sells the Grok models through its own API. Grok 4.6 is the current flagship and xAI tells developers to use it for everything outside audio, image and video, code included. It has a 500K context window and costs $2 in and $6 out per million tokens under 200K prompt tokens. Older Grok 4.20 and 4.3 models keep a 1M window at $1.25 in and $2.50 out, which is aggressive for that capability class.
Example models: Grok 4.6, Grok 4.20
Full xAI profileShould you choose Subconscious or xAI?
Subconscious
Choose Subconscious for
- Agent traces that would trip Grok's 2x rate past 200K
- Contexts beyond Grok 4.6's 500K window
- Open models on dedicated or on-prem hardware
xAI
Choose xAI for
- Agents that need live X posts and current events
- Cheap output tokens on prompts under 200K
- First-party image, video and audio APIs from the same vendor
Subconscious vs xAI at a glance
| Attribute | Subconscious | xAI |
|---|---|---|
| Model access | Open weights | Closed |
| Flagship models | GLM 5.3, DeepSeek V4.1 Flash | Grok 4.6, Grok 4.20, grok-build |
| Speed | 2x faster task completion | ~54 tok/s on Grok 4.6 |
| Price | 50–80% lower cost; billed on processed tokens | $2 in, $6 out (Grok 4.6); 2x past 200K |
| Customization | Marathon post-trained variants | Unknown |
| Deployment | Managed API, dedicated, on-prem | First-party API |
| Long context | 5M+ effective context | 500K (4.6), 1M (4.20, 4.3) |
Frequently asked questions
What is the difference between Subconscious and xAI?
Grok doubles the price of any request past 200K tokens. That is exactly where Subconscious is built to run, billing processed tokens with a 5M+ effective context.
When should I choose Subconscious over xAI?
Agent traces that would trip Grok's 2x rate past 200K; Contexts beyond Grok 4.6's 500K window; Open models on dedicated or on-prem hardware.
When should I choose xAI over Subconscious?
Agents that need live X posts and current events; Cheap output tokens on prompts under 200K; First-party image, video and audio APIs from the same vendor.
Is Subconscious or xAI cheaper?
Subconscious: 50–80% lower cost; billed on processed tokens. xAI: $2 in, $6 out (Grok 4.6); 2x past 200K. The cheaper choice depends on the model and workload.
Which has more context, Subconscious or xAI?
Subconscious: 5M+ effective context. xAI: 500K (4.6), 1M (4.20, 4.3).
Related comparisons
Run your longest agent traces on Subconscious
Point the OpenAI or Anthropic SDK, or the coding agent you already use, at Subconscious. Keep xAI for the work it does best and send the long runs to us.